diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -1,21882 +1,18039 @@ { "best_metric": null, "best_model_checkpoint": null, - "epoch": 0.999839871897518, + "epoch": 0.9996843204390374, "eval_steps": 500, - "global_step": 3122, + "global_step": 2573, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { - "epoch": 0.0003202562049639712, - "grad_norm": 0.5703125, - "learning_rate": 1.597444089456869e-07, - "loss": 2.5505, + "epoch": 0.0003885286904154829, + "grad_norm": 1.0859375, + "learning_rate": 7.751937984496125e-07, + "loss": 2.7508, "step": 1 }, { - "epoch": 0.0006405124099279424, - "grad_norm": 0.73828125, - "learning_rate": 3.194888178913738e-07, - "loss": 2.302, + "epoch": 0.0007770573808309658, + "grad_norm": 0.875, + "learning_rate": 1.550387596899225e-06, + "loss": 2.6944, "step": 2 }, { - "epoch": 0.0009607686148919135, - "grad_norm": 0.484375, - "learning_rate": 4.792332268370607e-07, - "loss": 2.0733, + "epoch": 0.0011655860712464486, + "grad_norm": 0.94921875, + "learning_rate": 2.325581395348837e-06, + "loss": 2.6677, "step": 3 }, { - "epoch": 0.0012810248198558848, - "grad_norm": 0.52734375, - "learning_rate": 6.389776357827476e-07, - "loss": 2.5665, + "epoch": 0.0015541147616619315, + "grad_norm": 0.89453125, + "learning_rate": 3.10077519379845e-06, + "loss": 2.6969, "step": 4 }, { - "epoch": 0.0016012810248198558, - "grad_norm": 0.578125, - "learning_rate": 7.987220447284345e-07, - "loss": 2.2262, + "epoch": 0.0019426434520774143, + "grad_norm": 0.81640625, + "learning_rate": 3.875968992248062e-06, + "loss": 2.6918, "step": 5 }, { - "epoch": 0.001921537229783827, - "grad_norm": 0.48828125, - "learning_rate": 9.584664536741213e-07, - "loss": 2.7195, + "epoch": 0.002331172142492897, + "grad_norm": 0.8203125, + "learning_rate": 4.651162790697674e-06, + "loss": 2.5546, "step": 6 }, { - "epoch": 0.0022417934347477983, - "grad_norm": 1.1484375, - "learning_rate": 1.1182108626198083e-06, - "loss": 2.3025, + "epoch": 0.00271970083290838, + "grad_norm": 0.984375, + "learning_rate": 5.426356589147287e-06, + "loss": 2.6413, "step": 7 }, { - "epoch": 0.0025620496397117695, - "grad_norm": 0.267578125, - "learning_rate": 1.2779552715654952e-06, - "loss": 1.7873, + "epoch": 0.003108229523323863, + "grad_norm": 0.9453125, + "learning_rate": 6.2015503875969e-06, + "loss": 2.7262, "step": 8 }, { - "epoch": 0.0028823058446757407, - "grad_norm": 0.38671875, - "learning_rate": 1.4376996805111822e-06, - "loss": 2.1716, + "epoch": 0.0034967582137393457, + "grad_norm": 0.85546875, + "learning_rate": 6.976744186046512e-06, + "loss": 2.7526, "step": 9 }, { - "epoch": 0.0032025620496397116, - "grad_norm": 0.6328125, - "learning_rate": 1.597444089456869e-06, - "loss": 2.1507, + "epoch": 0.0038852869041548286, + "grad_norm": 0.82421875, + "learning_rate": 7.751937984496124e-06, + "loss": 2.7219, "step": 10 }, { - "epoch": 0.003522818254603683, - "grad_norm": 0.77734375, - "learning_rate": 1.7571884984025559e-06, - "loss": 2.2224, + "epoch": 0.004273815594570312, + "grad_norm": 0.8359375, + "learning_rate": 8.527131782945736e-06, + "loss": 2.7289, "step": 11 }, { - "epoch": 0.003843074459567654, - "grad_norm": 0.5, - "learning_rate": 1.9169329073482426e-06, - "loss": 2.3008, + "epoch": 0.004662344284985794, + "grad_norm": 0.87109375, + "learning_rate": 9.302325581395349e-06, + "loss": 2.6555, "step": 12 }, { - "epoch": 0.004163330664531626, - "grad_norm": 0.4609375, - "learning_rate": 2.0766773162939296e-06, - "loss": 2.3001, + "epoch": 0.005050872975401278, + "grad_norm": 0.79296875, + "learning_rate": 1.0077519379844961e-05, + "loss": 2.6741, "step": 13 }, { - "epoch": 0.0044835868694955965, - "grad_norm": 0.52734375, - "learning_rate": 2.2364217252396165e-06, - "loss": 2.1583, + "epoch": 0.00543940166581676, + "grad_norm": 0.84375, + "learning_rate": 1.0852713178294575e-05, + "loss": 2.6498, "step": 14 }, { - "epoch": 0.004803843074459567, - "grad_norm": 0.60546875, - "learning_rate": 2.3961661341853035e-06, - "loss": 2.2752, + "epoch": 0.005827930356232243, + "grad_norm": 0.87109375, + "learning_rate": 1.1627906976744187e-05, + "loss": 2.7851, "step": 15 }, { - "epoch": 0.005124099279423539, - "grad_norm": 0.443359375, - "learning_rate": 2.5559105431309904e-06, - "loss": 1.6753, + "epoch": 0.006216459046647726, + "grad_norm": 0.828125, + "learning_rate": 1.24031007751938e-05, + "loss": 2.8024, "step": 16 }, { - "epoch": 0.00544435548438751, - "grad_norm": 1.4609375, - "learning_rate": 2.7156549520766774e-06, - "loss": 2.4427, + "epoch": 0.006604987737063209, + "grad_norm": 0.92578125, + "learning_rate": 1.3178294573643413e-05, + "loss": 2.7433, "step": 17 }, { - "epoch": 0.0057646116893514815, - "grad_norm": 1.0546875, - "learning_rate": 2.8753993610223644e-06, - "loss": 2.3853, + "epoch": 0.006993516427478691, + "grad_norm": 0.91796875, + "learning_rate": 1.3953488372093024e-05, + "loss": 2.687, "step": 18 }, { - "epoch": 0.006084867894315452, - "grad_norm": 0.51953125, - "learning_rate": 3.0351437699680513e-06, - "loss": 2.4479, + "epoch": 0.007382045117894175, + "grad_norm": 0.87109375, + "learning_rate": 1.4728682170542638e-05, + "loss": 2.7203, "step": 19 }, { - "epoch": 0.006405124099279423, - "grad_norm": 0.59375, - "learning_rate": 3.194888178913738e-06, - "loss": 1.9668, + "epoch": 0.007770573808309657, + "grad_norm": 0.96484375, + "learning_rate": 1.5503875968992248e-05, + "loss": 2.6724, "step": 20 }, { - "epoch": 0.006725380304243395, - "grad_norm": 1.2890625, - "learning_rate": 3.354632587859425e-06, - "loss": 2.2987, + "epoch": 0.00815910249872514, + "grad_norm": 0.8203125, + "learning_rate": 1.6279069767441862e-05, + "loss": 2.7522, "step": 21 }, { - "epoch": 0.007045636509207366, - "grad_norm": 0.48046875, - "learning_rate": 3.5143769968051118e-06, - "loss": 1.7833, + "epoch": 0.008547631189140623, + "grad_norm": 0.87890625, + "learning_rate": 1.7054263565891473e-05, + "loss": 2.692, "step": 22 }, { - "epoch": 0.007365892714171337, - "grad_norm": 0.494140625, - "learning_rate": 3.6741214057507987e-06, - "loss": 2.3454, + "epoch": 0.008936159879556106, + "grad_norm": 0.8125, + "learning_rate": 1.7829457364341087e-05, + "loss": 2.6184, "step": 23 }, { - "epoch": 0.007686148919135308, - "grad_norm": 0.609375, - "learning_rate": 3.833865814696485e-06, - "loss": 2.1669, + "epoch": 0.009324688569971588, + "grad_norm": 0.7890625, + "learning_rate": 1.8604651162790697e-05, + "loss": 2.6716, "step": 24 }, { - "epoch": 0.008006405124099279, - "grad_norm": 0.45703125, - "learning_rate": 3.993610223642173e-06, - "loss": 2.2589, + "epoch": 0.009713217260387071, + "grad_norm": 0.81640625, + "learning_rate": 1.937984496124031e-05, + "loss": 2.689, "step": 25 }, { - "epoch": 0.008326661329063251, - "grad_norm": 0.458984375, - "learning_rate": 4.153354632587859e-06, - "loss": 2.3922, + "epoch": 0.010101745950802555, + "grad_norm": 0.81640625, + "learning_rate": 2.0155038759689922e-05, + "loss": 2.623, "step": 26 }, { - "epoch": 0.008646917534027222, - "grad_norm": 0.5, - "learning_rate": 4.3130990415335465e-06, - "loss": 2.8444, + "epoch": 0.010490274641218038, + "grad_norm": 0.828125, + "learning_rate": 2.0930232558139536e-05, + "loss": 2.711, "step": 27 }, { - "epoch": 0.008967173738991193, - "grad_norm": 0.45703125, - "learning_rate": 4.472843450479233e-06, - "loss": 1.6251, + "epoch": 0.01087880333163352, + "grad_norm": 0.8125, + "learning_rate": 2.170542635658915e-05, + "loss": 2.7812, "step": 28 }, { - "epoch": 0.009287429943955164, - "grad_norm": 0.57421875, - "learning_rate": 4.6325878594249205e-06, - "loss": 1.4672, + "epoch": 0.011267332022049003, + "grad_norm": 0.8046875, + "learning_rate": 2.2480620155038764e-05, + "loss": 2.7191, "step": 29 }, { - "epoch": 0.009607686148919135, - "grad_norm": 0.5625, - "learning_rate": 4.792332268370607e-06, - "loss": 2.2324, + "epoch": 0.011655860712464486, + "grad_norm": 0.79296875, + "learning_rate": 2.3255813953488374e-05, + "loss": 2.7255, "step": 30 }, { - "epoch": 0.009927942353883107, - "grad_norm": 0.455078125, - "learning_rate": 4.952076677316294e-06, - "loss": 1.8752, + "epoch": 0.012044389402879968, + "grad_norm": 0.84375, + "learning_rate": 2.4031007751937988e-05, + "loss": 2.6695, "step": 31 }, { - "epoch": 0.010248198558847078, - "grad_norm": 1.921875, - "learning_rate": 5.111821086261981e-06, - "loss": 2.4251, + "epoch": 0.012432918093295452, + "grad_norm": 0.82421875, + "learning_rate": 2.48062015503876e-05, + "loss": 2.7645, "step": 32 }, { - "epoch": 0.010568454763811049, - "grad_norm": 0.412109375, - "learning_rate": 5.2715654952076674e-06, - "loss": 2.2702, + "epoch": 0.012821446783710935, + "grad_norm": 0.8046875, + "learning_rate": 2.5581395348837212e-05, + "loss": 2.7689, "step": 33 }, { - "epoch": 0.01088871096877502, - "grad_norm": 5.3125, - "learning_rate": 5.431309904153355e-06, - "loss": 2.2467, + "epoch": 0.013209975474126417, + "grad_norm": 0.8125, + "learning_rate": 2.6356589147286826e-05, + "loss": 2.7137, "step": 34 }, { - "epoch": 0.01120896717373899, - "grad_norm": 0.357421875, - "learning_rate": 5.591054313099041e-06, - "loss": 1.8738, + "epoch": 0.0135985041645419, + "grad_norm": 0.83984375, + "learning_rate": 2.7131782945736434e-05, + "loss": 2.6915, "step": 35 }, { - "epoch": 0.011529223378702963, - "grad_norm": 0.5234375, - "learning_rate": 5.750798722044729e-06, - "loss": 2.0557, + "epoch": 0.013987032854957383, + "grad_norm": 0.76953125, + "learning_rate": 2.7906976744186048e-05, + "loss": 2.6823, "step": 36 }, { - "epoch": 0.011849479583666934, - "grad_norm": 0.5390625, - "learning_rate": 5.910543130990415e-06, - "loss": 2.236, + "epoch": 0.014375561545372867, + "grad_norm": 0.796875, + "learning_rate": 2.868217054263566e-05, + "loss": 2.7233, "step": 37 }, { - "epoch": 0.012169735788630905, - "grad_norm": 0.984375, - "learning_rate": 6.070287539936103e-06, - "loss": 1.9037, + "epoch": 0.01476409023578835, + "grad_norm": 0.77734375, + "learning_rate": 2.9457364341085275e-05, + "loss": 2.7272, "step": 38 }, { - "epoch": 0.012489991993594875, - "grad_norm": 0.46484375, - "learning_rate": 6.230031948881789e-06, - "loss": 1.9012, + "epoch": 0.015152618926203832, + "grad_norm": 0.8125, + "learning_rate": 3.0232558139534883e-05, + "loss": 2.7218, "step": 39 }, { - "epoch": 0.012810248198558846, - "grad_norm": 2.1875, - "learning_rate": 6.389776357827476e-06, - "loss": 1.4725, + "epoch": 0.015541147616619315, + "grad_norm": 0.80078125, + "learning_rate": 3.1007751937984497e-05, + "loss": 2.7278, "step": 40 }, { - "epoch": 0.013130504403522819, - "grad_norm": 0.6796875, - "learning_rate": 6.549520766773164e-06, - "loss": 2.4334, + "epoch": 0.015929676307034797, + "grad_norm": 0.828125, + "learning_rate": 3.1782945736434114e-05, + "loss": 2.7345, "step": 41 }, { - "epoch": 0.01345076060848679, - "grad_norm": 0.48046875, - "learning_rate": 6.70926517571885e-06, - "loss": 2.2842, + "epoch": 0.01631820499745028, + "grad_norm": 0.8125, + "learning_rate": 3.2558139534883724e-05, + "loss": 2.6648, "step": 42 }, { - "epoch": 0.01377101681345076, - "grad_norm": 0.40625, - "learning_rate": 6.869009584664538e-06, - "loss": 1.993, + "epoch": 0.016706733687865762, + "grad_norm": 0.86328125, + "learning_rate": 3.3333333333333335e-05, + "loss": 2.6722, "step": 43 }, { - "epoch": 0.014091273018414731, - "grad_norm": 0.50390625, - "learning_rate": 7.0287539936102235e-06, - "loss": 1.7918, + "epoch": 0.017095262378281247, + "grad_norm": 0.7734375, + "learning_rate": 3.4108527131782945e-05, + "loss": 2.7215, "step": 44 }, { - "epoch": 0.014411529223378704, - "grad_norm": 0.4375, - "learning_rate": 7.188498402555911e-06, - "loss": 2.573, + "epoch": 0.017483791068696727, + "grad_norm": 0.80859375, + "learning_rate": 3.488372093023256e-05, + "loss": 2.7325, "step": 45 }, { - "epoch": 0.014731785428342675, - "grad_norm": 1.578125, - "learning_rate": 7.3482428115015974e-06, - "loss": 2.0233, + "epoch": 0.01787231975911221, + "grad_norm": 0.8359375, + "learning_rate": 3.565891472868217e-05, + "loss": 2.7248, "step": 46 }, { - "epoch": 0.015052041633306645, - "grad_norm": 0.47265625, - "learning_rate": 7.507987220447285e-06, - "loss": 2.1536, + "epoch": 0.018260848449527696, + "grad_norm": 0.81640625, + "learning_rate": 3.6434108527131784e-05, + "loss": 2.6442, "step": 47 }, { - "epoch": 0.015372297838270616, - "grad_norm": 0.5703125, - "learning_rate": 7.66773162939297e-06, - "loss": 2.4289, + "epoch": 0.018649377139943177, + "grad_norm": 0.75390625, + "learning_rate": 3.7209302325581394e-05, + "loss": 2.6807, "step": 48 }, { - "epoch": 0.015692554043234587, - "grad_norm": 0.412109375, - "learning_rate": 7.82747603833866e-06, - "loss": 1.7267, + "epoch": 0.01903790583035866, + "grad_norm": 0.75, + "learning_rate": 3.798449612403101e-05, + "loss": 2.5396, "step": 49 }, { - "epoch": 0.016012810248198558, - "grad_norm": 0.384765625, - "learning_rate": 7.987220447284345e-06, - "loss": 1.8542, + "epoch": 0.019426434520774142, + "grad_norm": 0.8359375, + "learning_rate": 3.875968992248062e-05, + "loss": 2.6923, "step": 50 }, { - "epoch": 0.01633306645316253, - "grad_norm": 0.494140625, - "learning_rate": 8.146964856230033e-06, - "loss": 1.6736, + "epoch": 0.019814963211189626, + "grad_norm": 0.796875, + "learning_rate": 3.953488372093023e-05, + "loss": 2.5531, "step": 51 }, { - "epoch": 0.016653322658126503, - "grad_norm": 0.59375, - "learning_rate": 8.306709265175718e-06, - "loss": 2.1386, + "epoch": 0.02020349190160511, + "grad_norm": 0.77734375, + "learning_rate": 4.0310077519379843e-05, + "loss": 2.6413, "step": 52 }, { - "epoch": 0.016973578863090474, - "grad_norm": 0.45703125, - "learning_rate": 8.466453674121406e-06, - "loss": 2.3592, + "epoch": 0.02059202059202059, + "grad_norm": 0.79296875, + "learning_rate": 4.108527131782946e-05, + "loss": 2.6221, "step": 53 }, { - "epoch": 0.017293835068054444, - "grad_norm": 0.546875, - "learning_rate": 8.626198083067093e-06, - "loss": 1.732, + "epoch": 0.020980549282436076, + "grad_norm": 0.78515625, + "learning_rate": 4.186046511627907e-05, + "loss": 2.6596, "step": 54 }, { - "epoch": 0.017614091273018415, - "grad_norm": 0.77734375, - "learning_rate": 8.78594249201278e-06, - "loss": 1.5725, + "epoch": 0.021369077972851556, + "grad_norm": 0.76953125, + "learning_rate": 4.263565891472868e-05, + "loss": 2.6862, "step": 55 }, { - "epoch": 0.017934347477982386, - "grad_norm": 0.3984375, - "learning_rate": 8.945686900958466e-06, - "loss": 2.4561, + "epoch": 0.02175760666326704, + "grad_norm": 0.78125, + "learning_rate": 4.34108527131783e-05, + "loss": 2.7227, "step": 56 }, { - "epoch": 0.018254603682946357, - "grad_norm": 0.734375, - "learning_rate": 9.105431309904154e-06, - "loss": 1.8126, + "epoch": 0.022146135353682525, + "grad_norm": 0.7890625, + "learning_rate": 4.418604651162791e-05, + "loss": 2.7295, "step": 57 }, { - "epoch": 0.018574859887910328, - "grad_norm": 0.474609375, - "learning_rate": 9.265175718849841e-06, - "loss": 2.1243, + "epoch": 0.022534664044098006, + "grad_norm": 0.7578125, + "learning_rate": 4.496124031007753e-05, + "loss": 2.7031, "step": 58 }, { - "epoch": 0.0188951160928743, - "grad_norm": 0.54296875, - "learning_rate": 9.424920127795528e-06, - "loss": 2.3236, + "epoch": 0.02292319273451349, + "grad_norm": 0.87890625, + "learning_rate": 4.573643410852713e-05, + "loss": 2.7625, "step": 59 }, { - "epoch": 0.01921537229783827, - "grad_norm": 0.5234375, - "learning_rate": 9.584664536741214e-06, - "loss": 2.3305, + "epoch": 0.02331172142492897, + "grad_norm": 0.76953125, + "learning_rate": 4.651162790697675e-05, + "loss": 2.7127, "step": 60 }, { - "epoch": 0.01953562850280224, - "grad_norm": 0.49609375, - "learning_rate": 9.744408945686901e-06, - "loss": 2.1348, + "epoch": 0.023700250115344455, + "grad_norm": 0.80078125, + "learning_rate": 4.728682170542636e-05, + "loss": 2.6414, "step": 61 }, { - "epoch": 0.019855884707766214, - "grad_norm": 0.314453125, - "learning_rate": 9.904153354632589e-06, - "loss": 1.91, + "epoch": 0.024088778805759936, + "grad_norm": 0.77734375, + "learning_rate": 4.8062015503875976e-05, + "loss": 2.6865, "step": 62 }, { - "epoch": 0.020176140912730185, - "grad_norm": 0.53515625, - "learning_rate": 1.0063897763578276e-05, - "loss": 2.4788, + "epoch": 0.02447730749617542, + "grad_norm": 0.82421875, + "learning_rate": 4.883720930232558e-05, + "loss": 2.6609, "step": 63 }, { - "epoch": 0.020496397117694156, - "grad_norm": 0.84765625, - "learning_rate": 1.0223642172523962e-05, - "loss": 2.4574, + "epoch": 0.024865836186590905, + "grad_norm": 0.859375, + "learning_rate": 4.96124031007752e-05, + "loss": 2.7198, "step": 64 }, { - "epoch": 0.020816653322658127, - "grad_norm": 0.8203125, - "learning_rate": 1.038338658146965e-05, - "loss": 2.3022, + "epoch": 0.025254364877006386, + "grad_norm": 0.8125, + "learning_rate": 5.038759689922481e-05, + "loss": 2.6678, "step": 65 }, { - "epoch": 0.021136909527622098, - "grad_norm": 0.470703125, - "learning_rate": 1.0543130990415335e-05, - "loss": 1.4106, + "epoch": 0.02564289356742187, + "grad_norm": 0.7578125, + "learning_rate": 5.1162790697674425e-05, + "loss": 2.7294, "step": 66 }, { - "epoch": 0.02145716573258607, - "grad_norm": 0.52734375, - "learning_rate": 1.0702875399361024e-05, - "loss": 2.4533, + "epoch": 0.02603142225783735, + "grad_norm": 0.69140625, + "learning_rate": 5.1937984496124036e-05, + "loss": 2.6087, "step": 67 }, { - "epoch": 0.02177742193755004, - "grad_norm": 0.66015625, - "learning_rate": 1.086261980830671e-05, - "loss": 2.6732, + "epoch": 0.026419950948252835, + "grad_norm": 0.75390625, + "learning_rate": 5.271317829457365e-05, + "loss": 2.6123, "step": 68 }, { - "epoch": 0.02209767814251401, - "grad_norm": 0.609375, - "learning_rate": 1.1022364217252397e-05, - "loss": 2.09, + "epoch": 0.02680847963866832, + "grad_norm": 0.76171875, + "learning_rate": 5.348837209302326e-05, + "loss": 2.6227, "step": 69 }, { - "epoch": 0.02241793434747798, - "grad_norm": 0.451171875, - "learning_rate": 1.1182108626198083e-05, - "loss": 2.4639, + "epoch": 0.0271970083290838, + "grad_norm": 0.80859375, + "learning_rate": 5.426356589147287e-05, + "loss": 2.6388, "step": 70 }, { - "epoch": 0.022738190552441955, - "grad_norm": 0.5859375, - "learning_rate": 1.134185303514377e-05, - "loss": 2.195, + "epoch": 0.027585537019499284, + "grad_norm": 0.79296875, + "learning_rate": 5.503875968992248e-05, + "loss": 2.6372, "step": 71 }, { - "epoch": 0.023058446757405926, - "grad_norm": 0.55078125, - "learning_rate": 1.1501597444089457e-05, - "loss": 2.0553, + "epoch": 0.027974065709914765, + "grad_norm": 0.7578125, + "learning_rate": 5.5813953488372095e-05, + "loss": 2.6499, "step": 72 }, { - "epoch": 0.023378702962369897, - "grad_norm": 0.46484375, - "learning_rate": 1.1661341853035145e-05, - "loss": 1.6604, + "epoch": 0.02836259440033025, + "grad_norm": 0.79296875, + "learning_rate": 5.6589147286821706e-05, + "loss": 2.6805, "step": 73 }, { - "epoch": 0.023698959167333868, - "grad_norm": 0.48046875, - "learning_rate": 1.182108626198083e-05, - "loss": 2.1982, + "epoch": 0.028751123090745734, + "grad_norm": 0.78125, + "learning_rate": 5.736434108527132e-05, + "loss": 2.646, "step": 74 }, { - "epoch": 0.02401921537229784, - "grad_norm": 0.45703125, - "learning_rate": 1.1980830670926518e-05, - "loss": 2.2353, + "epoch": 0.029139651781161215, + "grad_norm": 0.83984375, + "learning_rate": 5.8139534883720933e-05, + "loss": 2.6443, "step": 75 }, { - "epoch": 0.02433947157726181, - "grad_norm": 0.4609375, - "learning_rate": 1.2140575079872205e-05, - "loss": 2.0923, + "epoch": 0.0295281804715767, + "grad_norm": 0.8515625, + "learning_rate": 5.891472868217055e-05, + "loss": 2.7152, "step": 76 }, { - "epoch": 0.02465972778222578, - "grad_norm": 0.5078125, - "learning_rate": 1.2300319488817893e-05, - "loss": 2.379, + "epoch": 0.02991670916199218, + "grad_norm": 0.84375, + "learning_rate": 5.9689922480620155e-05, + "loss": 2.6011, "step": 77 }, { - "epoch": 0.02497998398718975, - "grad_norm": 1.0625, - "learning_rate": 1.2460063897763578e-05, - "loss": 1.7649, + "epoch": 0.030305237852407664, + "grad_norm": 0.84765625, + "learning_rate": 6.0465116279069765e-05, + "loss": 2.5998, "step": 78 }, { - "epoch": 0.02530024019215372, - "grad_norm": 0.64453125, - "learning_rate": 1.2619808306709266e-05, - "loss": 2.475, + "epoch": 0.030693766542823145, + "grad_norm": 0.8671875, + "learning_rate": 6.124031007751938e-05, + "loss": 2.6207, "step": 79 }, { - "epoch": 0.025620496397117692, - "grad_norm": 0.453125, - "learning_rate": 1.2779552715654951e-05, - "loss": 2.4901, + "epoch": 0.03108229523323863, + "grad_norm": 0.75390625, + "learning_rate": 6.201550387596899e-05, + "loss": 2.616, "step": 80 }, { - "epoch": 0.025940752602081667, - "grad_norm": 0.458984375, - "learning_rate": 1.2939297124600639e-05, - "loss": 2.4741, + "epoch": 0.03147082392365411, + "grad_norm": 0.796875, + "learning_rate": 6.27906976744186e-05, + "loss": 2.6722, "step": 81 }, { - "epoch": 0.026261008807045638, - "grad_norm": 0.37890625, - "learning_rate": 1.3099041533546328e-05, - "loss": 2.1992, + "epoch": 0.031859352614069594, + "grad_norm": 0.78125, + "learning_rate": 6.356589147286823e-05, + "loss": 2.5915, "step": 82 }, { - "epoch": 0.02658126501200961, - "grad_norm": 0.65234375, - "learning_rate": 1.3258785942492014e-05, - "loss": 2.6322, + "epoch": 0.032247881304485075, + "grad_norm": 0.80078125, + "learning_rate": 6.434108527131784e-05, + "loss": 2.6935, "step": 83 }, { - "epoch": 0.02690152121697358, - "grad_norm": 0.78515625, - "learning_rate": 1.34185303514377e-05, - "loss": 2.2932, + "epoch": 0.03263640999490056, + "grad_norm": 0.80859375, + "learning_rate": 6.511627906976745e-05, + "loss": 2.5366, "step": 84 }, { - "epoch": 0.02722177742193755, - "grad_norm": 0.462890625, - "learning_rate": 1.3578274760383387e-05, - "loss": 1.828, + "epoch": 0.033024938685316044, + "grad_norm": 0.70703125, + "learning_rate": 6.589147286821705e-05, + "loss": 2.5981, "step": 85 }, { - "epoch": 0.02754203362690152, - "grad_norm": 0.490234375, - "learning_rate": 1.3738019169329076e-05, - "loss": 2.0083, + "epoch": 0.033413467375731525, + "grad_norm": 0.765625, + "learning_rate": 6.666666666666667e-05, + "loss": 2.598, "step": 86 }, { - "epoch": 0.02786228983186549, - "grad_norm": 0.37109375, - "learning_rate": 1.3897763578274761e-05, - "loss": 2.0055, + "epoch": 0.03380199606614701, + "grad_norm": 0.87109375, + "learning_rate": 6.744186046511628e-05, + "loss": 2.7113, "step": 87 }, { - "epoch": 0.028182546036829462, - "grad_norm": 2.890625, - "learning_rate": 1.4057507987220447e-05, - "loss": 2.3489, + "epoch": 0.03419052475656249, + "grad_norm": 0.83984375, + "learning_rate": 6.821705426356589e-05, + "loss": 2.6221, "step": 88 }, { - "epoch": 0.028502802241793433, - "grad_norm": 0.5390625, - "learning_rate": 1.4217252396166134e-05, - "loss": 2.0319, + "epoch": 0.034579053446977974, + "grad_norm": 0.7890625, + "learning_rate": 6.89922480620155e-05, + "loss": 2.5659, "step": 89 }, { - "epoch": 0.028823058446757407, - "grad_norm": 0.6015625, - "learning_rate": 1.4376996805111822e-05, - "loss": 1.8418, + "epoch": 0.034967582137393455, + "grad_norm": 0.83203125, + "learning_rate": 6.976744186046513e-05, + "loss": 2.6699, "step": 90 }, { - "epoch": 0.02914331465172138, - "grad_norm": 0.6640625, - "learning_rate": 1.453674121405751e-05, - "loss": 2.1933, + "epoch": 0.03535611082780894, + "grad_norm": 0.77734375, + "learning_rate": 7.054263565891474e-05, + "loss": 2.6132, "step": 91 }, { - "epoch": 0.02946357085668535, - "grad_norm": 0.51953125, - "learning_rate": 1.4696485623003195e-05, - "loss": 2.238, + "epoch": 0.03574463951822442, + "grad_norm": 0.78515625, + "learning_rate": 7.131782945736435e-05, + "loss": 2.536, "step": 92 }, { - "epoch": 0.02978382706164932, - "grad_norm": 0.494140625, - "learning_rate": 1.485623003194888e-05, - "loss": 1.8154, + "epoch": 0.036133168208639904, + "grad_norm": 0.89453125, + "learning_rate": 7.209302325581396e-05, + "loss": 2.6658, "step": 93 }, { - "epoch": 0.03010408326661329, - "grad_norm": 0.375, - "learning_rate": 1.501597444089457e-05, - "loss": 2.0664, + "epoch": 0.03652169689905539, + "grad_norm": 0.8359375, + "learning_rate": 7.286821705426357e-05, + "loss": 2.5863, "step": 94 }, { - "epoch": 0.03042433947157726, - "grad_norm": 0.56640625, - "learning_rate": 1.5175718849840257e-05, - "loss": 2.4674, + "epoch": 0.03691022558947087, + "grad_norm": 0.734375, + "learning_rate": 7.364341085271318e-05, + "loss": 2.5564, "step": 95 }, { - "epoch": 0.030744595676541232, - "grad_norm": 0.404296875, - "learning_rate": 1.533546325878594e-05, - "loss": 2.5225, + "epoch": 0.037298754279886354, + "grad_norm": 0.8203125, + "learning_rate": 7.441860465116279e-05, + "loss": 2.627, "step": 96 }, { - "epoch": 0.031064851881505203, - "grad_norm": 0.796875, - "learning_rate": 1.549520766773163e-05, - "loss": 2.5732, + "epoch": 0.03768728297030184, + "grad_norm": 0.94921875, + "learning_rate": 7.519379844961241e-05, + "loss": 2.6368, "step": 97 }, { - "epoch": 0.031385108086469174, - "grad_norm": 0.484375, - "learning_rate": 1.565495207667732e-05, - "loss": 2.4263, + "epoch": 0.03807581166071732, + "grad_norm": 0.890625, + "learning_rate": 7.596899224806202e-05, + "loss": 2.5973, "step": 98 }, { - "epoch": 0.03170536429143315, - "grad_norm": 0.45703125, - "learning_rate": 1.5814696485623005e-05, - "loss": 1.8861, + "epoch": 0.0384643403511328, + "grad_norm": 1.96875, + "learning_rate": 7.674418604651163e-05, + "loss": 2.5859, "step": 99 }, { - "epoch": 0.032025620496397116, - "grad_norm": 0.44921875, - "learning_rate": 1.597444089456869e-05, - "loss": 2.187, + "epoch": 0.038852869041548284, + "grad_norm": 0.76953125, + "learning_rate": 7.751937984496124e-05, + "loss": 2.6385, "step": 100 }, { - "epoch": 0.03234587670136109, - "grad_norm": 0.58203125, - "learning_rate": 1.6134185303514376e-05, - "loss": 2.4841, + "epoch": 0.03924139773196377, + "grad_norm": 0.85546875, + "learning_rate": 7.829457364341086e-05, + "loss": 2.6456, "step": 101 }, { - "epoch": 0.03266613290632506, - "grad_norm": 0.88671875, - "learning_rate": 1.6293929712460065e-05, - "loss": 2.1184, + "epoch": 0.03962992642237925, + "grad_norm": 0.75, + "learning_rate": 7.906976744186047e-05, + "loss": 2.5322, "step": 102 }, { - "epoch": 0.03298638911128903, - "grad_norm": 0.47265625, - "learning_rate": 1.645367412140575e-05, - "loss": 2.5642, + "epoch": 0.04001845511279473, + "grad_norm": 0.78125, + "learning_rate": 7.984496124031008e-05, + "loss": 2.5148, "step": 103 }, { - "epoch": 0.033306645316253006, - "grad_norm": 0.318359375, - "learning_rate": 1.6613418530351437e-05, - "loss": 2.2907, + "epoch": 0.04040698380321022, + "grad_norm": 0.78125, + "learning_rate": 8.062015503875969e-05, + "loss": 2.5003, "step": 104 }, { - "epoch": 0.03362690152121697, - "grad_norm": 0.3828125, - "learning_rate": 1.6773162939297126e-05, - "loss": 2.4367, + "epoch": 0.0407955124936257, + "grad_norm": 0.87890625, + "learning_rate": 8.139534883720931e-05, + "loss": 2.6288, "step": 105 }, { - "epoch": 0.03394715772618095, - "grad_norm": 0.4140625, - "learning_rate": 1.693290734824281e-05, - "loss": 2.5302, + "epoch": 0.04118404118404118, + "grad_norm": 0.73828125, + "learning_rate": 8.217054263565892e-05, + "loss": 2.659, "step": 106 }, { - "epoch": 0.034267413931144915, - "grad_norm": 0.5703125, - "learning_rate": 1.70926517571885e-05, - "loss": 2.1865, + "epoch": 0.041572569874456664, + "grad_norm": 0.83984375, + "learning_rate": 8.294573643410853e-05, + "loss": 2.5254, "step": 107 }, { - "epoch": 0.03458767013610889, - "grad_norm": 0.55078125, - "learning_rate": 1.7252396166134186e-05, - "loss": 2.1695, + "epoch": 0.04196109856487215, + "grad_norm": 0.79296875, + "learning_rate": 8.372093023255814e-05, + "loss": 2.5423, "step": 108 }, { - "epoch": 0.034907926341072856, - "grad_norm": 0.52734375, - "learning_rate": 1.7412140575079875e-05, - "loss": 2.3653, + "epoch": 0.04234962725528763, + "grad_norm": 0.7890625, + "learning_rate": 8.449612403100775e-05, + "loss": 2.549, "step": 109 }, { - "epoch": 0.03522818254603683, - "grad_norm": 0.396484375, - "learning_rate": 1.757188498402556e-05, - "loss": 1.9492, + "epoch": 0.04273815594570311, + "grad_norm": 0.796875, + "learning_rate": 8.527131782945736e-05, + "loss": 2.6419, "step": 110 }, { - "epoch": 0.0355484387510008, - "grad_norm": 0.388671875, - "learning_rate": 1.7731629392971247e-05, - "loss": 2.7231, + "epoch": 0.0431266846361186, + "grad_norm": 0.8671875, + "learning_rate": 8.604651162790697e-05, + "loss": 2.5485, "step": 111 }, { - "epoch": 0.03586869495596477, - "grad_norm": 0.40234375, - "learning_rate": 1.7891373801916932e-05, - "loss": 2.2126, + "epoch": 0.04351521332653408, + "grad_norm": 0.8515625, + "learning_rate": 8.68217054263566e-05, + "loss": 2.5752, "step": 112 }, { - "epoch": 0.03618895116092874, - "grad_norm": 0.416015625, - "learning_rate": 1.805111821086262e-05, - "loss": 2.4667, + "epoch": 0.04390374201694956, + "grad_norm": 0.84375, + "learning_rate": 8.759689922480621e-05, + "loss": 2.5693, "step": 113 }, { - "epoch": 0.036509207365892714, - "grad_norm": 0.64453125, - "learning_rate": 1.8210862619808307e-05, - "loss": 2.3994, + "epoch": 0.04429227070736505, + "grad_norm": 0.828125, + "learning_rate": 8.837209302325582e-05, + "loss": 2.5739, "step": 114 }, { - "epoch": 0.03682946357085669, - "grad_norm": 0.5859375, - "learning_rate": 1.8370607028753993e-05, - "loss": 2.2501, + "epoch": 0.04468079939778053, + "grad_norm": 0.83203125, + "learning_rate": 8.914728682170543e-05, + "loss": 2.5961, "step": 115 }, { - "epoch": 0.037149719775820655, - "grad_norm": 1.703125, - "learning_rate": 1.8530351437699682e-05, - "loss": 2.5277, + "epoch": 0.04506932808819601, + "grad_norm": 0.78125, + "learning_rate": 8.992248062015505e-05, + "loss": 2.5351, "step": 116 }, { - "epoch": 0.03746997598078463, - "grad_norm": 0.39453125, - "learning_rate": 1.869009584664537e-05, - "loss": 2.1283, + "epoch": 0.04545785677861149, + "grad_norm": 0.76953125, + "learning_rate": 9.069767441860465e-05, + "loss": 2.5145, "step": 117 }, { - "epoch": 0.0377902321857486, - "grad_norm": 1.6328125, - "learning_rate": 1.8849840255591057e-05, - "loss": 2.5671, + "epoch": 0.04584638546902698, + "grad_norm": 0.796875, + "learning_rate": 9.147286821705426e-05, + "loss": 2.5982, "step": 118 }, { - "epoch": 0.03811048839071257, - "grad_norm": 1.796875, - "learning_rate": 1.9009584664536742e-05, - "loss": 1.7323, + "epoch": 0.04623491415944246, + "grad_norm": 0.75390625, + "learning_rate": 9.224806201550387e-05, + "loss": 2.5882, "step": 119 }, { - "epoch": 0.03843074459567654, - "grad_norm": 0.486328125, - "learning_rate": 1.9169329073482428e-05, - "loss": 2.4496, + "epoch": 0.04662344284985794, + "grad_norm": 0.84765625, + "learning_rate": 9.30232558139535e-05, + "loss": 2.6582, "step": 120 }, { - "epoch": 0.03875100080064051, - "grad_norm": 0.6171875, - "learning_rate": 1.9329073482428117e-05, - "loss": 2.3247, + "epoch": 0.04701197154027343, + "grad_norm": 0.83203125, + "learning_rate": 9.379844961240311e-05, + "loss": 2.5619, "step": 121 }, { - "epoch": 0.03907125700560448, - "grad_norm": 0.318359375, - "learning_rate": 1.9488817891373803e-05, - "loss": 2.044, + "epoch": 0.04740050023068891, + "grad_norm": 0.90234375, + "learning_rate": 9.457364341085272e-05, + "loss": 2.588, "step": 122 }, { - "epoch": 0.039391513210568455, - "grad_norm": 0.55859375, - "learning_rate": 1.964856230031949e-05, - "loss": 2.2774, + "epoch": 0.04778902892110439, + "grad_norm": 0.7734375, + "learning_rate": 9.534883720930233e-05, + "loss": 2.5661, "step": 123 }, { - "epoch": 0.03971176941553243, - "grad_norm": 0.765625, - "learning_rate": 1.9808306709265177e-05, - "loss": 2.6735, + "epoch": 0.04817755761151987, + "grad_norm": 0.83984375, + "learning_rate": 9.612403100775195e-05, + "loss": 2.5462, "step": 124 }, { - "epoch": 0.040032025620496396, - "grad_norm": 0.498046875, - "learning_rate": 1.9968051118210863e-05, - "loss": 2.5676, + "epoch": 0.04856608630193536, + "grad_norm": 0.8203125, + "learning_rate": 9.689922480620155e-05, + "loss": 2.5859, "step": 125 }, { - "epoch": 0.04035228182546037, - "grad_norm": 0.41015625, - "learning_rate": 2.0127795527156552e-05, - "loss": 2.132, + "epoch": 0.04895461499235084, + "grad_norm": 0.89453125, + "learning_rate": 9.767441860465116e-05, + "loss": 2.6039, "step": 126 }, { - "epoch": 0.04067253803042434, - "grad_norm": 0.4609375, - "learning_rate": 2.0287539936102238e-05, - "loss": 2.6514, + "epoch": 0.04934314368276632, + "grad_norm": 0.84765625, + "learning_rate": 9.844961240310078e-05, + "loss": 2.6323, "step": 127 }, { - "epoch": 0.04099279423538831, - "grad_norm": 0.65234375, - "learning_rate": 2.0447284345047924e-05, - "loss": 2.6641, + "epoch": 0.04973167237318181, + "grad_norm": 0.75390625, + "learning_rate": 9.92248062015504e-05, + "loss": 2.5727, "step": 128 }, { - "epoch": 0.04131305044035228, - "grad_norm": 0.494140625, - "learning_rate": 2.0607028753993613e-05, - "loss": 2.442, + "epoch": 0.05012020106359729, + "grad_norm": 0.81640625, + "learning_rate": 0.0001, + "loss": 2.604, "step": 129 }, { - "epoch": 0.041633306645316254, - "grad_norm": 0.3359375, - "learning_rate": 2.07667731629393e-05, - "loss": 1.4356, + "epoch": 0.05050872975401277, + "grad_norm": 0.7890625, + "learning_rate": 0.00010077519379844962, + "loss": 2.5792, "step": 130 }, { - "epoch": 0.04195356285028022, - "grad_norm": 0.34765625, - "learning_rate": 2.0926517571884984e-05, - "loss": 1.8376, + "epoch": 0.05089725844442826, + "grad_norm": 0.7265625, + "learning_rate": 0.00010155038759689923, + "loss": 2.5262, "step": 131 }, { - "epoch": 0.042273819055244195, - "grad_norm": 0.3671875, - "learning_rate": 2.108626198083067e-05, - "loss": 2.4422, + "epoch": 0.05128578713484374, + "grad_norm": 0.82421875, + "learning_rate": 0.00010232558139534885, + "loss": 2.6018, "step": 132 }, { - "epoch": 0.04259407526020817, - "grad_norm": 0.578125, - "learning_rate": 2.124600638977636e-05, - "loss": 2.1663, + "epoch": 0.05167431582525922, + "grad_norm": 0.875, + "learning_rate": 0.00010310077519379846, + "loss": 2.5635, "step": 133 }, { - "epoch": 0.04291433146517214, - "grad_norm": 0.6171875, - "learning_rate": 2.1405750798722048e-05, - "loss": 2.5161, + "epoch": 0.0520628445156747, + "grad_norm": 0.77734375, + "learning_rate": 0.00010387596899224807, + "loss": 2.5924, "step": 134 }, { - "epoch": 0.04323458767013611, - "grad_norm": 0.6328125, - "learning_rate": 2.1565495207667734e-05, - "loss": 2.295, + "epoch": 0.05245137320609019, + "grad_norm": 0.84375, + "learning_rate": 0.00010465116279069768, + "loss": 2.6196, "step": 135 }, { - "epoch": 0.04355484387510008, - "grad_norm": 0.48046875, - "learning_rate": 2.172523961661342e-05, - "loss": 2.6238, + "epoch": 0.05283990189650567, + "grad_norm": 0.8359375, + "learning_rate": 0.0001054263565891473, + "loss": 2.5211, "step": 136 }, { - "epoch": 0.04387510008006405, - "grad_norm": 1.9609375, - "learning_rate": 2.188498402555911e-05, - "loss": 2.0364, + "epoch": 0.05322843058692115, + "grad_norm": 0.78125, + "learning_rate": 0.00010620155038759692, + "loss": 2.5739, "step": 137 }, { - "epoch": 0.04419535628502802, - "grad_norm": 0.77734375, - "learning_rate": 2.2044728434504794e-05, - "loss": 2.7412, + "epoch": 0.05361695927733664, + "grad_norm": 0.7890625, + "learning_rate": 0.00010697674418604651, + "loss": 2.5201, "step": 138 }, { - "epoch": 0.044515612489991994, - "grad_norm": 1.0703125, - "learning_rate": 2.220447284345048e-05, - "loss": 1.8156, + "epoch": 0.05400548796775212, + "grad_norm": 0.7578125, + "learning_rate": 0.00010775193798449612, + "loss": 2.5191, "step": 139 }, { - "epoch": 0.04483586869495596, - "grad_norm": 0.91796875, - "learning_rate": 2.2364217252396165e-05, - "loss": 2.3815, + "epoch": 0.0543940166581676, + "grad_norm": 0.88671875, + "learning_rate": 0.00010852713178294573, + "loss": 2.5712, "step": 140 }, { - "epoch": 0.045156124899919936, - "grad_norm": 0.439453125, - "learning_rate": 2.2523961661341854e-05, - "loss": 2.6271, + "epoch": 0.05478254534858308, + "grad_norm": 0.80859375, + "learning_rate": 0.00010930232558139534, + "loss": 2.5579, "step": 141 }, { - "epoch": 0.04547638110488391, - "grad_norm": 1.7421875, - "learning_rate": 2.268370607028754e-05, - "loss": 1.6873, + "epoch": 0.05517107403899857, + "grad_norm": 0.86328125, + "learning_rate": 0.00011007751937984496, + "loss": 2.5769, "step": 142 }, { - "epoch": 0.04579663730984788, - "grad_norm": 0.431640625, - "learning_rate": 2.284345047923323e-05, - "loss": 2.3417, + "epoch": 0.05555960272941405, + "grad_norm": 0.81640625, + "learning_rate": 0.00011085271317829458, + "loss": 2.5025, "step": 143 }, { - "epoch": 0.04611689351481185, - "grad_norm": 0.4609375, - "learning_rate": 2.3003194888178915e-05, - "loss": 2.1776, + "epoch": 0.05594813141982953, + "grad_norm": 0.80859375, + "learning_rate": 0.00011162790697674419, + "loss": 2.4641, "step": 144 }, { - "epoch": 0.04643714971977582, - "grad_norm": 0.498046875, - "learning_rate": 2.3162939297124604e-05, - "loss": 2.2616, + "epoch": 0.05633666011024502, + "grad_norm": 0.78125, + "learning_rate": 0.0001124031007751938, + "loss": 2.4801, "step": 145 }, { - "epoch": 0.046757405924739794, - "grad_norm": 0.60546875, - "learning_rate": 2.332268370607029e-05, - "loss": 2.5723, + "epoch": 0.0567251888006605, + "grad_norm": 0.8515625, + "learning_rate": 0.00011317829457364341, + "loss": 2.5683, "step": 146 }, { - "epoch": 0.04707766212970376, - "grad_norm": 0.470703125, - "learning_rate": 2.3482428115015975e-05, - "loss": 2.4548, + "epoch": 0.05711371749107598, + "grad_norm": 0.8046875, + "learning_rate": 0.00011395348837209304, + "loss": 2.5542, "step": 147 }, { - "epoch": 0.047397918334667735, - "grad_norm": 0.9375, - "learning_rate": 2.364217252396166e-05, - "loss": 2.4736, + "epoch": 0.05750224618149147, + "grad_norm": 0.796875, + "learning_rate": 0.00011472868217054265, + "loss": 2.4648, "step": 148 }, { - "epoch": 0.0477181745396317, - "grad_norm": 0.482421875, - "learning_rate": 2.380191693290735e-05, - "loss": 2.386, + "epoch": 0.05789077487190695, + "grad_norm": 0.8203125, + "learning_rate": 0.00011550387596899226, + "loss": 2.548, "step": 149 }, { - "epoch": 0.04803843074459568, - "grad_norm": 0.486328125, - "learning_rate": 2.3961661341853036e-05, - "loss": 2.644, + "epoch": 0.05827930356232243, + "grad_norm": 0.80078125, + "learning_rate": 0.00011627906976744187, + "loss": 2.6214, "step": 150 }, { - "epoch": 0.04835868694955965, - "grad_norm": 0.58203125, - "learning_rate": 2.412140575079872e-05, - "loss": 2.1945, + "epoch": 0.05866783225273791, + "grad_norm": 0.859375, + "learning_rate": 0.00011705426356589149, + "loss": 2.5642, "step": 151 }, { - "epoch": 0.04867894315452362, - "grad_norm": 0.66796875, - "learning_rate": 2.428115015974441e-05, - "loss": 2.2383, + "epoch": 0.0590563609431534, + "grad_norm": 0.76171875, + "learning_rate": 0.0001178294573643411, + "loss": 2.5584, "step": 152 }, { - "epoch": 0.04899919935948759, - "grad_norm": 0.4296875, - "learning_rate": 2.44408945686901e-05, - "loss": 2.5643, + "epoch": 0.05944488963356888, + "grad_norm": 0.80859375, + "learning_rate": 0.00011860465116279071, + "loss": 2.5286, "step": 153 }, { - "epoch": 0.04931945556445156, - "grad_norm": 0.431640625, - "learning_rate": 2.4600638977635785e-05, - "loss": 1.6623, + "epoch": 0.05983341832398436, + "grad_norm": 0.875, + "learning_rate": 0.00011937984496124031, + "loss": 2.6014, "step": 154 }, { - "epoch": 0.049639711769415534, - "grad_norm": 0.7890625, - "learning_rate": 2.476038338658147e-05, - "loss": 2.1212, + "epoch": 0.06022194701439985, + "grad_norm": 0.8671875, + "learning_rate": 0.00012015503875968992, + "loss": 2.5721, "step": 155 }, { - "epoch": 0.0499599679743795, - "grad_norm": 0.453125, - "learning_rate": 2.4920127795527157e-05, - "loss": 2.0499, + "epoch": 0.06061047570481533, + "grad_norm": 0.8359375, + "learning_rate": 0.00012093023255813953, + "loss": 2.5731, "step": 156 }, { - "epoch": 0.050280224179343476, - "grad_norm": 0.58984375, - "learning_rate": 2.5079872204472842e-05, - "loss": 2.004, + "epoch": 0.06099900439523081, + "grad_norm": 0.79296875, + "learning_rate": 0.00012170542635658914, + "loss": 2.4835, "step": 157 }, { - "epoch": 0.05060048038430744, - "grad_norm": 0.50390625, - "learning_rate": 2.523961661341853e-05, - "loss": 2.4388, + "epoch": 0.06138753308564629, + "grad_norm": 0.7890625, + "learning_rate": 0.00012248062015503876, + "loss": 2.5875, "step": 158 }, { - "epoch": 0.05092073658927142, - "grad_norm": 0.431640625, - "learning_rate": 2.539936102236422e-05, - "loss": 2.2752, + "epoch": 0.06177606177606178, + "grad_norm": 0.77734375, + "learning_rate": 0.00012325581395348836, + "loss": 2.5394, "step": 159 }, { - "epoch": 0.051240992794235385, - "grad_norm": 0.470703125, - "learning_rate": 2.5559105431309903e-05, - "loss": 2.1717, + "epoch": 0.06216459046647726, + "grad_norm": 0.75390625, + "learning_rate": 0.00012403100775193799, + "loss": 2.5477, "step": 160 }, { - "epoch": 0.05156124899919936, - "grad_norm": 0.46875, - "learning_rate": 2.5718849840255592e-05, - "loss": 2.3856, + "epoch": 0.06255311915689274, + "grad_norm": 0.8359375, + "learning_rate": 0.0001248062015503876, + "loss": 2.4915, "step": 161 }, { - "epoch": 0.05188150520416333, - "grad_norm": 0.96484375, - "learning_rate": 2.5878594249201278e-05, - "loss": 1.7901, + "epoch": 0.06294164784730823, + "grad_norm": 0.859375, + "learning_rate": 0.0001255813953488372, + "loss": 2.5073, "step": 162 }, { - "epoch": 0.0522017614091273, - "grad_norm": 0.62890625, - "learning_rate": 2.6038338658146967e-05, - "loss": 2.1485, + "epoch": 0.0633301765377237, + "grad_norm": 0.9140625, + "learning_rate": 0.00012635658914728683, + "loss": 2.5843, "step": 163 }, { - "epoch": 0.052522017614091275, - "grad_norm": 1.046875, - "learning_rate": 2.6198083067092656e-05, - "loss": 2.4288, + "epoch": 0.06371870522813919, + "grad_norm": 0.83984375, + "learning_rate": 0.00012713178294573646, + "loss": 2.5689, "step": 164 }, { - "epoch": 0.05284227381905524, - "grad_norm": 0.333984375, - "learning_rate": 2.6357827476038338e-05, - "loss": 2.4144, + "epoch": 0.06410723391855468, + "grad_norm": 0.796875, + "learning_rate": 0.00012790697674418605, + "loss": 2.5615, "step": 165 }, { - "epoch": 0.05316253002401922, - "grad_norm": 0.390625, - "learning_rate": 2.6517571884984027e-05, - "loss": 2.4021, + "epoch": 0.06449576260897015, + "grad_norm": 0.8359375, + "learning_rate": 0.00012868217054263568, + "loss": 2.519, "step": 166 }, { - "epoch": 0.053482786228983184, - "grad_norm": 0.439453125, - "learning_rate": 2.6677316293929716e-05, - "loss": 2.2037, + "epoch": 0.06488429129938564, + "grad_norm": 0.90625, + "learning_rate": 0.00012945736434108527, + "loss": 2.596, "step": 167 }, { - "epoch": 0.05380304243394716, - "grad_norm": 0.63671875, - "learning_rate": 2.68370607028754e-05, - "loss": 2.33, + "epoch": 0.06527281998980113, + "grad_norm": 0.8203125, + "learning_rate": 0.0001302325581395349, + "loss": 2.5081, "step": 168 }, { - "epoch": 0.054123298638911126, - "grad_norm": 0.7421875, - "learning_rate": 2.6996805111821088e-05, - "loss": 2.3497, + "epoch": 0.0656613486802166, + "grad_norm": 0.80078125, + "learning_rate": 0.00013100775193798452, + "loss": 2.5528, "step": 169 }, { - "epoch": 0.0544435548438751, - "grad_norm": 0.4765625, - "learning_rate": 2.7156549520766773e-05, - "loss": 2.0505, + "epoch": 0.06604987737063209, + "grad_norm": 0.828125, + "learning_rate": 0.0001317829457364341, + "loss": 2.5174, "step": 170 }, { - "epoch": 0.054763811048839074, - "grad_norm": 0.478515625, - "learning_rate": 2.7316293929712462e-05, - "loss": 2.5688, + "epoch": 0.06643840606104758, + "grad_norm": 0.765625, + "learning_rate": 0.00013255813953488372, + "loss": 2.5159, "step": 171 }, { - "epoch": 0.05508406725380304, - "grad_norm": 0.57421875, - "learning_rate": 2.747603833865815e-05, - "loss": 2.5961, + "epoch": 0.06682693475146305, + "grad_norm": 0.78515625, + "learning_rate": 0.00013333333333333334, + "loss": 2.5556, "step": 172 }, { - "epoch": 0.055404323458767016, - "grad_norm": 0.490234375, - "learning_rate": 2.7635782747603834e-05, - "loss": 1.9069, + "epoch": 0.06721546344187854, + "grad_norm": 0.8125, + "learning_rate": 0.00013410852713178294, + "loss": 2.5453, "step": 173 }, { - "epoch": 0.05572457966373098, - "grad_norm": 0.498046875, - "learning_rate": 2.7795527156549523e-05, - "loss": 2.3172, + "epoch": 0.06760399213229402, + "grad_norm": 0.8046875, + "learning_rate": 0.00013488372093023256, + "loss": 2.5447, "step": 174 }, { - "epoch": 0.05604483586869496, - "grad_norm": 0.67578125, - "learning_rate": 2.7955271565495212e-05, - "loss": 2.5545, + "epoch": 0.0679925208227095, + "grad_norm": 0.79296875, + "learning_rate": 0.00013565891472868218, + "loss": 2.5426, "step": 175 }, { - "epoch": 0.056365092073658925, - "grad_norm": 0.75390625, - "learning_rate": 2.8115015974440894e-05, - "loss": 2.248, + "epoch": 0.06838104951312499, + "grad_norm": 0.765625, + "learning_rate": 0.00013643410852713178, + "loss": 2.5165, "step": 176 }, { - "epoch": 0.0566853482786229, - "grad_norm": 0.486328125, - "learning_rate": 2.8274760383386583e-05, - "loss": 2.4285, + "epoch": 0.06876957820354047, + "grad_norm": 0.85546875, + "learning_rate": 0.0001372093023255814, + "loss": 2.4734, "step": 177 }, { - "epoch": 0.057005604483586866, - "grad_norm": 0.373046875, - "learning_rate": 2.843450479233227e-05, - "loss": 1.4118, + "epoch": 0.06915810689395595, + "grad_norm": 0.78515625, + "learning_rate": 0.000137984496124031, + "loss": 2.4351, "step": 178 }, { - "epoch": 0.05732586068855084, - "grad_norm": 0.6328125, - "learning_rate": 2.8594249201277955e-05, - "loss": 1.8937, + "epoch": 0.06954663558437144, + "grad_norm": 0.83203125, + "learning_rate": 0.00013875968992248063, + "loss": 2.5237, "step": 179 }, { - "epoch": 0.057646116893514815, - "grad_norm": 0.482421875, - "learning_rate": 2.8753993610223644e-05, - "loss": 1.7478, + "epoch": 0.06993516427478691, + "grad_norm": 0.76953125, + "learning_rate": 0.00013953488372093025, + "loss": 2.5439, "step": 180 }, { - "epoch": 0.05796637309847878, - "grad_norm": 0.55859375, - "learning_rate": 2.891373801916933e-05, - "loss": 1.8904, + "epoch": 0.0703236929652024, + "grad_norm": 0.8203125, + "learning_rate": 0.00014031007751937985, + "loss": 2.4856, "step": 181 }, { - "epoch": 0.05828662930344276, - "grad_norm": 0.455078125, - "learning_rate": 2.907348242811502e-05, - "loss": 2.6784, + "epoch": 0.07071222165561789, + "grad_norm": 0.9140625, + "learning_rate": 0.00014108527131782947, + "loss": 2.4619, "step": 182 }, { - "epoch": 0.058606885508406724, - "grad_norm": 0.515625, - "learning_rate": 2.9233226837060707e-05, - "loss": 2.4389, + "epoch": 0.07110075034603336, + "grad_norm": 0.8046875, + "learning_rate": 0.0001418604651162791, + "loss": 2.5004, "step": 183 }, { - "epoch": 0.0589271417133707, - "grad_norm": 0.5703125, - "learning_rate": 2.939297124600639e-05, - "loss": 2.5659, + "epoch": 0.07148927903644885, + "grad_norm": 0.828125, + "learning_rate": 0.0001426356589147287, + "loss": 2.4954, "step": 184 }, { - "epoch": 0.059247397918334666, - "grad_norm": 0.498046875, - "learning_rate": 2.955271565495208e-05, - "loss": 2.0976, + "epoch": 0.07187780772686433, + "grad_norm": 0.8515625, + "learning_rate": 0.00014341085271317832, + "loss": 2.5676, "step": 185 }, { - "epoch": 0.05956765412329864, - "grad_norm": 1.546875, - "learning_rate": 2.971246006389776e-05, - "loss": 2.5468, + "epoch": 0.07226633641727981, + "grad_norm": 0.86328125, + "learning_rate": 0.00014418604651162791, + "loss": 2.5853, "step": 186 }, { - "epoch": 0.05988791032826261, - "grad_norm": 0.455078125, - "learning_rate": 2.987220447284345e-05, - "loss": 2.3513, + "epoch": 0.0726548651076953, + "grad_norm": 0.76953125, + "learning_rate": 0.0001449612403100775, + "loss": 2.4586, "step": 187 }, { - "epoch": 0.06020816653322658, - "grad_norm": 0.5234375, - "learning_rate": 3.003194888178914e-05, - "loss": 1.9096, + "epoch": 0.07304339379811078, + "grad_norm": 0.9609375, + "learning_rate": 0.00014573643410852714, + "loss": 2.4955, "step": 188 }, { - "epoch": 0.060528422738190556, - "grad_norm": 0.44921875, - "learning_rate": 3.0191693290734825e-05, - "loss": 1.768, + "epoch": 0.07343192248852626, + "grad_norm": 0.9296875, + "learning_rate": 0.00014651162790697673, + "loss": 2.5761, "step": 189 }, { - "epoch": 0.06084867894315452, - "grad_norm": 0.404296875, - "learning_rate": 3.0351437699680514e-05, - "loss": 1.7802, + "epoch": 0.07382045117894175, + "grad_norm": 0.796875, + "learning_rate": 0.00014728682170542636, + "loss": 2.5159, "step": 190 }, { - "epoch": 0.0611689351481185, - "grad_norm": 0.46484375, - "learning_rate": 3.0511182108626203e-05, - "loss": 1.9335, + "epoch": 0.07420897986935723, + "grad_norm": 0.8828125, + "learning_rate": 0.00014806201550387598, + "loss": 2.4549, "step": 191 }, { - "epoch": 0.061489191353082465, - "grad_norm": 0.439453125, - "learning_rate": 3.067092651757188e-05, - "loss": 2.3403, + "epoch": 0.07459750855977271, + "grad_norm": 0.88671875, + "learning_rate": 0.00014883720930232558, + "loss": 2.4911, "step": 192 }, { - "epoch": 0.06180944755804644, - "grad_norm": 0.51171875, - "learning_rate": 3.083067092651757e-05, - "loss": 2.1596, + "epoch": 0.0749860372501882, + "grad_norm": 0.8515625, + "learning_rate": 0.0001496124031007752, + "loss": 2.5391, "step": 193 }, { - "epoch": 0.062129703763010406, - "grad_norm": 0.390625, - "learning_rate": 3.099041533546326e-05, - "loss": 2.6126, + "epoch": 0.07537456594060368, + "grad_norm": 0.875, + "learning_rate": 0.00015038759689922483, + "loss": 2.5419, "step": 194 }, { - "epoch": 0.06244995996797438, - "grad_norm": 0.322265625, - "learning_rate": 3.115015974440895e-05, - "loss": 1.772, + "epoch": 0.07576309463101916, + "grad_norm": 0.91796875, + "learning_rate": 0.00015116279069767442, + "loss": 2.5355, "step": 195 }, { - "epoch": 0.06277021617293835, - "grad_norm": 0.63671875, - "learning_rate": 3.130990415335464e-05, - "loss": 2.694, + "epoch": 0.07615162332143464, + "grad_norm": 0.8671875, + "learning_rate": 0.00015193798449612405, + "loss": 2.5481, "step": 196 }, { - "epoch": 0.06309047237790232, - "grad_norm": 0.484375, - "learning_rate": 3.146964856230032e-05, - "loss": 2.045, + "epoch": 0.07654015201185012, + "grad_norm": 0.8359375, + "learning_rate": 0.00015271317829457364, + "loss": 2.5098, "step": 197 }, { - "epoch": 0.0634107285828663, - "grad_norm": 1.828125, - "learning_rate": 3.162939297124601e-05, - "loss": 1.7398, + "epoch": 0.0769286807022656, + "grad_norm": 0.81640625, + "learning_rate": 0.00015348837209302327, + "loss": 2.47, "step": 198 }, { - "epoch": 0.06373098478783026, - "grad_norm": 0.59375, - "learning_rate": 3.17891373801917e-05, - "loss": 2.344, + "epoch": 0.0773172093926811, + "grad_norm": 0.875, + "learning_rate": 0.0001542635658914729, + "loss": 2.5496, "step": 199 }, { - "epoch": 0.06405124099279423, - "grad_norm": 0.5546875, - "learning_rate": 3.194888178913738e-05, - "loss": 2.1837, + "epoch": 0.07770573808309657, + "grad_norm": 0.921875, + "learning_rate": 0.0001550387596899225, + "loss": 2.579, "step": 200 }, { - "epoch": 0.06437149719775821, - "grad_norm": 1.171875, - "learning_rate": 3.210862619808307e-05, - "loss": 2.0782, + "epoch": 0.07809426677351206, + "grad_norm": 0.89453125, + "learning_rate": 0.0001558139534883721, + "loss": 2.538, "step": 201 }, { - "epoch": 0.06469175340272218, - "grad_norm": 0.44921875, - "learning_rate": 3.226837060702875e-05, - "loss": 2.2655, + "epoch": 0.07848279546392754, + "grad_norm": 0.83203125, + "learning_rate": 0.0001565891472868217, + "loss": 2.4561, "step": 202 }, { - "epoch": 0.06501200960768615, - "grad_norm": 0.462890625, - "learning_rate": 3.242811501597444e-05, - "loss": 2.55, + "epoch": 0.07887132415434302, + "grad_norm": 0.88671875, + "learning_rate": 0.0001573643410852713, + "loss": 2.4939, "step": 203 }, { - "epoch": 0.06533226581265011, - "grad_norm": 0.59765625, - "learning_rate": 3.258785942492013e-05, - "loss": 2.2424, + "epoch": 0.0792598528447585, + "grad_norm": 0.93359375, + "learning_rate": 0.00015813953488372093, + "loss": 2.5002, "step": 204 }, { - "epoch": 0.0656525220176141, - "grad_norm": 0.486328125, - "learning_rate": 3.274760383386581e-05, - "loss": 2.5929, + "epoch": 0.07964838153517399, + "grad_norm": 0.81640625, + "learning_rate": 0.00015891472868217056, + "loss": 2.4702, "step": 205 }, { - "epoch": 0.06597277822257806, - "grad_norm": 0.400390625, - "learning_rate": 3.29073482428115e-05, - "loss": 2.2174, + "epoch": 0.08003691022558947, + "grad_norm": 0.8125, + "learning_rate": 0.00015968992248062015, + "loss": 2.4821, "step": 206 }, { - "epoch": 0.06629303442754203, - "grad_norm": 0.5078125, - "learning_rate": 3.306709265175719e-05, - "loss": 2.685, + "epoch": 0.08042543891600495, + "grad_norm": 0.94140625, + "learning_rate": 0.00016046511627906978, + "loss": 2.5318, "step": 207 }, { - "epoch": 0.06661329063250601, - "grad_norm": 0.486328125, - "learning_rate": 3.322683706070287e-05, - "loss": 2.242, + "epoch": 0.08081396760642044, + "grad_norm": 0.90234375, + "learning_rate": 0.00016124031007751937, + "loss": 2.4792, "step": 208 }, { - "epoch": 0.06693354683746998, - "grad_norm": 0.4375, - "learning_rate": 3.338658146964856e-05, - "loss": 1.9813, + "epoch": 0.08120249629683592, + "grad_norm": 0.78515625, + "learning_rate": 0.000162015503875969, + "loss": 2.432, "step": 209 }, { - "epoch": 0.06725380304243395, - "grad_norm": 0.5390625, - "learning_rate": 3.354632587859425e-05, - "loss": 2.0788, + "epoch": 0.0815910249872514, + "grad_norm": 0.90625, + "learning_rate": 0.00016279069767441862, + "loss": 2.5145, "step": 210 }, { - "epoch": 0.06757405924739791, - "grad_norm": 0.478515625, - "learning_rate": 3.3706070287539934e-05, - "loss": 2.2187, + "epoch": 0.08197955367766689, + "grad_norm": 0.84765625, + "learning_rate": 0.00016356589147286822, + "loss": 2.4505, "step": 211 }, { - "epoch": 0.0678943154523619, - "grad_norm": 0.4765625, - "learning_rate": 3.386581469648562e-05, - "loss": 2.2772, + "epoch": 0.08236808236808237, + "grad_norm": 0.94921875, + "learning_rate": 0.00016434108527131784, + "loss": 2.5375, "step": 212 }, { - "epoch": 0.06821457165732586, - "grad_norm": 0.373046875, - "learning_rate": 3.402555910543131e-05, - "loss": 1.4853, + "epoch": 0.08275661105849785, + "grad_norm": 0.91796875, + "learning_rate": 0.00016511627906976747, + "loss": 2.4642, "step": 213 }, { - "epoch": 0.06853482786228983, - "grad_norm": 0.53515625, - "learning_rate": 3.4185303514377e-05, - "loss": 2.0633, + "epoch": 0.08314513974891333, + "grad_norm": 0.9140625, + "learning_rate": 0.00016589147286821706, + "loss": 2.5279, "step": 214 }, { - "epoch": 0.0688550840672538, - "grad_norm": 0.6953125, - "learning_rate": 3.434504792332269e-05, - "loss": 2.3552, + "epoch": 0.08353366843932881, + "grad_norm": 0.8828125, + "learning_rate": 0.0001666666666666667, + "loss": 2.5083, "step": 215 }, { - "epoch": 0.06917534027221778, - "grad_norm": 0.498046875, - "learning_rate": 3.450479233226837e-05, - "loss": 2.7011, + "epoch": 0.0839221971297443, + "grad_norm": 0.98046875, + "learning_rate": 0.00016744186046511629, + "loss": 2.5472, "step": 216 }, { - "epoch": 0.06949559647718175, - "grad_norm": 0.439453125, - "learning_rate": 3.466453674121406e-05, - "loss": 2.0144, + "epoch": 0.08431072582015978, + "grad_norm": 0.9296875, + "learning_rate": 0.0001682170542635659, + "loss": 2.5635, "step": 217 }, { - "epoch": 0.06981585268214571, - "grad_norm": 0.640625, - "learning_rate": 3.482428115015975e-05, - "loss": 2.293, + "epoch": 0.08469925451057526, + "grad_norm": 0.9296875, + "learning_rate": 0.0001689922480620155, + "loss": 2.4878, "step": 218 }, { - "epoch": 0.0701361088871097, - "grad_norm": 0.376953125, - "learning_rate": 3.498402555910543e-05, - "loss": 2.2046, + "epoch": 0.08508778320099075, + "grad_norm": 0.9296875, + "learning_rate": 0.0001697674418604651, + "loss": 2.5143, "step": 219 }, { - "epoch": 0.07045636509207366, - "grad_norm": 1.0390625, - "learning_rate": 3.514376996805112e-05, - "loss": 2.0654, + "epoch": 0.08547631189140623, + "grad_norm": 0.94921875, + "learning_rate": 0.00017054263565891473, + "loss": 2.5344, "step": 220 }, { - "epoch": 0.07077662129703763, - "grad_norm": 1.1484375, - "learning_rate": 3.5303514376996804e-05, - "loss": 2.3501, + "epoch": 0.08586484058182171, + "grad_norm": 0.83984375, + "learning_rate": 0.00017131782945736435, + "loss": 2.5229, "step": 221 }, { - "epoch": 0.0710968775020016, - "grad_norm": 0.51171875, - "learning_rate": 3.546325878594249e-05, - "loss": 2.3258, + "epoch": 0.0862533692722372, + "grad_norm": 0.9375, + "learning_rate": 0.00017209302325581395, + "loss": 2.5182, "step": 222 }, { - "epoch": 0.07141713370696558, - "grad_norm": 0.41796875, - "learning_rate": 3.562300319488818e-05, - "loss": 2.3782, + "epoch": 0.08664189796265268, + "grad_norm": 0.91015625, + "learning_rate": 0.00017286821705426357, + "loss": 2.5788, "step": 223 }, { - "epoch": 0.07173738991192954, - "grad_norm": 0.5078125, - "learning_rate": 3.5782747603833865e-05, - "loss": 2.2688, + "epoch": 0.08703042665306816, + "grad_norm": 0.8984375, + "learning_rate": 0.0001736434108527132, + "loss": 2.4586, "step": 224 }, { - "epoch": 0.07205764611689351, - "grad_norm": 0.67578125, - "learning_rate": 3.5942492012779554e-05, - "loss": 2.2505, + "epoch": 0.08741895534348365, + "grad_norm": 0.984375, + "learning_rate": 0.0001744186046511628, + "loss": 2.4999, "step": 225 }, { - "epoch": 0.07237790232185748, - "grad_norm": 1.0859375, - "learning_rate": 3.610223642172524e-05, - "loss": 2.4121, + "epoch": 0.08780748403389912, + "grad_norm": 0.84765625, + "learning_rate": 0.00017519379844961242, + "loss": 2.4922, "step": 226 }, { - "epoch": 0.07269815852682146, - "grad_norm": 0.55078125, - "learning_rate": 3.6261980830670925e-05, - "loss": 2.4804, + "epoch": 0.08819601272431461, + "grad_norm": 1.15625, + "learning_rate": 0.00017596899224806201, + "loss": 2.5033, "step": 227 }, { - "epoch": 0.07301841473178543, - "grad_norm": 0.6640625, - "learning_rate": 3.6421725239616614e-05, - "loss": 2.4248, + "epoch": 0.0885845414147301, + "grad_norm": 0.953125, + "learning_rate": 0.00017674418604651164, + "loss": 2.5631, "step": 228 }, { - "epoch": 0.0733386709367494, - "grad_norm": 0.91015625, - "learning_rate": 3.65814696485623e-05, - "loss": 2.2826, + "epoch": 0.08897307010514557, + "grad_norm": 1.1640625, + "learning_rate": 0.00017751937984496126, + "loss": 2.5307, "step": 229 }, { - "epoch": 0.07365892714171338, - "grad_norm": 0.392578125, - "learning_rate": 3.6741214057507985e-05, - "loss": 1.8805, + "epoch": 0.08936159879556106, + "grad_norm": 0.9375, + "learning_rate": 0.00017829457364341086, + "loss": 2.5699, "step": 230 }, { - "epoch": 0.07397918334667734, - "grad_norm": 1.265625, - "learning_rate": 3.6900958466453675e-05, - "loss": 1.9896, + "epoch": 0.08975012748597654, + "grad_norm": 1.0546875, + "learning_rate": 0.00017906976744186048, + "loss": 2.5328, "step": 231 }, { - "epoch": 0.07429943955164131, - "grad_norm": 1.515625, - "learning_rate": 3.7060702875399364e-05, - "loss": 2.6494, + "epoch": 0.09013865617639202, + "grad_norm": 0.9765625, + "learning_rate": 0.0001798449612403101, + "loss": 2.4664, "step": 232 }, { - "epoch": 0.07461969575660528, - "grad_norm": 0.5390625, - "learning_rate": 3.722044728434505e-05, - "loss": 1.5742, + "epoch": 0.09052718486680751, + "grad_norm": 0.9609375, + "learning_rate": 0.0001806201550387597, + "loss": 2.4832, "step": 233 }, { - "epoch": 0.07493995196156926, - "grad_norm": 0.353515625, - "learning_rate": 3.738019169329074e-05, - "loss": 1.9401, + "epoch": 0.09091571355722299, + "grad_norm": 0.921875, + "learning_rate": 0.0001813953488372093, + "loss": 2.5115, "step": 234 }, { - "epoch": 0.07526020816653323, - "grad_norm": 0.46875, - "learning_rate": 3.7539936102236424e-05, - "loss": 2.5195, + "epoch": 0.09130424224763847, + "grad_norm": 0.921875, + "learning_rate": 0.00018217054263565893, + "loss": 2.4564, "step": 235 }, { - "epoch": 0.0755804643714972, - "grad_norm": 0.4375, - "learning_rate": 3.769968051118211e-05, - "loss": 2.1442, + "epoch": 0.09169277093805396, + "grad_norm": 0.9375, + "learning_rate": 0.00018294573643410852, + "loss": 2.4124, "step": 236 }, { - "epoch": 0.07590072057646118, - "grad_norm": 2.015625, - "learning_rate": 3.7859424920127795e-05, - "loss": 2.2005, + "epoch": 0.09208129962846943, + "grad_norm": 0.85546875, + "learning_rate": 0.00018372093023255815, + "loss": 2.4601, "step": 237 }, { - "epoch": 0.07622097678142514, - "grad_norm": 0.4609375, - "learning_rate": 3.8019169329073485e-05, - "loss": 2.5538, + "epoch": 0.09246982831888492, + "grad_norm": 0.9296875, + "learning_rate": 0.00018449612403100774, + "loss": 2.5447, "step": 238 }, { - "epoch": 0.07654123298638911, - "grad_norm": 0.6875, - "learning_rate": 3.8178913738019174e-05, - "loss": 2.22, + "epoch": 0.09285835700930041, + "grad_norm": 0.90234375, + "learning_rate": 0.00018527131782945737, + "loss": 2.4984, "step": 239 }, { - "epoch": 0.07686148919135308, - "grad_norm": 0.54296875, - "learning_rate": 3.8338658146964856e-05, - "loss": 2.4174, + "epoch": 0.09324688569971588, + "grad_norm": 0.90234375, + "learning_rate": 0.000186046511627907, + "loss": 2.4448, "step": 240 }, { - "epoch": 0.07718174539631706, - "grad_norm": 0.478515625, - "learning_rate": 3.8498402555910545e-05, - "loss": 2.3724, + "epoch": 0.09363541439013137, + "grad_norm": 0.90234375, + "learning_rate": 0.0001868217054263566, + "loss": 2.501, "step": 241 }, { - "epoch": 0.07750200160128103, - "grad_norm": 0.59375, - "learning_rate": 3.8658146964856234e-05, - "loss": 2.1111, + "epoch": 0.09402394308054686, + "grad_norm": 0.9375, + "learning_rate": 0.00018759689922480621, + "loss": 2.4614, "step": 242 }, { - "epoch": 0.077822257806245, - "grad_norm": 0.6171875, - "learning_rate": 3.8817891373801916e-05, - "loss": 2.2114, + "epoch": 0.09441247177096233, + "grad_norm": 0.8984375, + "learning_rate": 0.00018837209302325584, + "loss": 2.5024, "step": 243 }, { - "epoch": 0.07814251401120896, - "grad_norm": 0.5, - "learning_rate": 3.8977635782747605e-05, - "loss": 2.718, + "epoch": 0.09480100046137782, + "grad_norm": 0.87890625, + "learning_rate": 0.00018914728682170543, + "loss": 2.4858, "step": 244 }, { - "epoch": 0.07846277021617294, - "grad_norm": 0.5859375, - "learning_rate": 3.913738019169329e-05, - "loss": 2.1236, + "epoch": 0.09518952915179331, + "grad_norm": 1.015625, + "learning_rate": 0.00018992248062015506, + "loss": 2.55, "step": 245 }, { - "epoch": 0.07878302642113691, - "grad_norm": 0.6171875, - "learning_rate": 3.929712460063898e-05, - "loss": 2.3918, + "epoch": 0.09557805784220878, + "grad_norm": 0.9375, + "learning_rate": 0.00019069767441860466, + "loss": 2.5271, "step": 246 }, { - "epoch": 0.07910328262610088, - "grad_norm": 0.474609375, - "learning_rate": 3.9456869009584666e-05, - "loss": 1.8762, + "epoch": 0.09596658653262427, + "grad_norm": 0.875, + "learning_rate": 0.00019147286821705428, + "loss": 2.4523, "step": 247 }, { - "epoch": 0.07942353883106486, - "grad_norm": 0.61328125, - "learning_rate": 3.9616613418530355e-05, - "loss": 2.2969, + "epoch": 0.09635511522303974, + "grad_norm": 0.9140625, + "learning_rate": 0.0001922480620155039, + "loss": 2.4744, "step": 248 }, { - "epoch": 0.07974379503602883, - "grad_norm": 0.4375, - "learning_rate": 3.9776357827476044e-05, - "loss": 2.1578, + "epoch": 0.09674364391345523, + "grad_norm": 0.87890625, + "learning_rate": 0.0001930232558139535, + "loss": 2.5274, "step": 249 }, { - "epoch": 0.08006405124099279, - "grad_norm": 0.46484375, - "learning_rate": 3.9936102236421726e-05, - "loss": 2.5525, + "epoch": 0.09713217260387072, + "grad_norm": 0.90234375, + "learning_rate": 0.0001937984496124031, + "loss": 2.5121, "step": 250 }, { - "epoch": 0.08038430744595676, - "grad_norm": 0.4921875, - "learning_rate": 4.0095846645367415e-05, - "loss": 2.4644, + "epoch": 0.0975207012942862, + "grad_norm": 0.8515625, + "learning_rate": 0.00019457364341085272, + "loss": 2.482, "step": 251 }, { - "epoch": 0.08070456365092074, - "grad_norm": 0.400390625, - "learning_rate": 4.0255591054313104e-05, - "loss": 1.2111, + "epoch": 0.09790922998470168, + "grad_norm": 0.84375, + "learning_rate": 0.00019534883720930232, + "loss": 2.5032, "step": 252 }, { - "epoch": 0.08102481985588471, - "grad_norm": 0.451171875, - "learning_rate": 4.041533546325879e-05, - "loss": 2.522, + "epoch": 0.09829775867511717, + "grad_norm": 0.90234375, + "learning_rate": 0.00019612403100775194, + "loss": 2.5395, "step": 253 }, { - "epoch": 0.08134507606084868, - "grad_norm": 0.41796875, - "learning_rate": 4.0575079872204476e-05, - "loss": 1.5268, + "epoch": 0.09868628736553264, + "grad_norm": 0.8984375, + "learning_rate": 0.00019689922480620157, + "loss": 2.485, "step": 254 }, { - "epoch": 0.08166533226581266, - "grad_norm": 0.408203125, - "learning_rate": 4.0734824281150165e-05, - "loss": 2.331, + "epoch": 0.09907481605594813, + "grad_norm": 0.8984375, + "learning_rate": 0.00019767441860465116, + "loss": 2.4981, "step": 255 }, { - "epoch": 0.08198558847077662, - "grad_norm": 0.8828125, - "learning_rate": 4.089456869009585e-05, - "loss": 2.2693, + "epoch": 0.09946334474636362, + "grad_norm": 0.94921875, + "learning_rate": 0.0001984496124031008, + "loss": 2.5291, "step": 256 }, { - "epoch": 0.08230584467574059, - "grad_norm": 0.58203125, - "learning_rate": 4.1054313099041536e-05, - "loss": 2.0041, + "epoch": 0.09985187343677909, + "grad_norm": 0.88671875, + "learning_rate": 0.00019922480620155039, + "loss": 2.4618, "step": 257 }, { - "epoch": 0.08262610088070456, - "grad_norm": 0.30078125, - "learning_rate": 4.1214057507987225e-05, - "loss": 1.814, + "epoch": 0.10024040212719458, + "grad_norm": 0.87109375, + "learning_rate": 0.0002, + "loss": 2.4954, "step": 258 }, { - "epoch": 0.08294635708566854, - "grad_norm": 0.59375, - "learning_rate": 4.137380191693291e-05, - "loss": 2.1807, + "epoch": 0.10062893081761007, + "grad_norm": 0.828125, + "learning_rate": 0.000199999907919495, + "loss": 2.5519, "step": 259 }, { - "epoch": 0.08326661329063251, - "grad_norm": 0.431640625, - "learning_rate": 4.15335463258786e-05, - "loss": 1.8303, + "epoch": 0.10101745950802554, + "grad_norm": 0.9296875, + "learning_rate": 0.00019999963167814967, + "loss": 2.5409, "step": 260 }, { - "epoch": 0.08358686949559647, - "grad_norm": 0.46484375, - "learning_rate": 4.169329073482428e-05, - "loss": 2.0817, + "epoch": 0.10140598819844103, + "grad_norm": 0.96875, + "learning_rate": 0.0001999991712764726, + "loss": 2.4918, "step": 261 }, { - "epoch": 0.08390712570056044, - "grad_norm": 0.384765625, - "learning_rate": 4.185303514376997e-05, - "loss": 2.572, + "epoch": 0.10179451688885652, + "grad_norm": 0.96875, + "learning_rate": 0.0001999985267153118, + "loss": 2.5241, "step": 262 }, { - "epoch": 0.08422738190552442, - "grad_norm": 0.5234375, - "learning_rate": 4.201277955271566e-05, - "loss": 2.4427, + "epoch": 0.10218304557927199, + "grad_norm": 0.91015625, + "learning_rate": 0.00019999769799585422, + "loss": 2.5064, "step": 263 }, { - "epoch": 0.08454763811048839, - "grad_norm": 0.470703125, - "learning_rate": 4.217252396166134e-05, - "loss": 1.6031, + "epoch": 0.10257157426968748, + "grad_norm": 0.87109375, + "learning_rate": 0.00019999668511962607, + "loss": 2.5119, "step": 264 }, { - "epoch": 0.08486789431545236, - "grad_norm": 0.388671875, - "learning_rate": 4.233226837060703e-05, - "loss": 2.3756, + "epoch": 0.10296010296010295, + "grad_norm": 1.0, + "learning_rate": 0.00019999548808849268, + "loss": 2.4937, "step": 265 }, { - "epoch": 0.08518815052041634, - "grad_norm": 0.75, - "learning_rate": 4.249201277955272e-05, - "loss": 2.5675, + "epoch": 0.10334863165051844, + "grad_norm": 1.0546875, + "learning_rate": 0.00019999410690465852, + "loss": 2.5707, "step": 266 }, { - "epoch": 0.0855084067253803, - "grad_norm": 0.6171875, - "learning_rate": 4.265175718849841e-05, - "loss": 2.3992, + "epoch": 0.10373716034093393, + "grad_norm": 0.94140625, + "learning_rate": 0.00019999254157066716, + "loss": 2.458, "step": 267 }, { - "epoch": 0.08582866293034427, - "grad_norm": 2.296875, - "learning_rate": 4.2811501597444096e-05, - "loss": 1.8396, + "epoch": 0.1041256890313494, + "grad_norm": 1.0625, + "learning_rate": 0.00019999079208940135, + "loss": 2.5633, "step": 268 }, { - "epoch": 0.08614891913530824, - "grad_norm": 0.625, - "learning_rate": 4.297124600638978e-05, - "loss": 1.8471, + "epoch": 0.10451421772176489, + "grad_norm": 0.91796875, + "learning_rate": 0.00019998885846408295, + "loss": 2.5172, "step": 269 }, { - "epoch": 0.08646917534027222, - "grad_norm": 0.6171875, - "learning_rate": 4.313099041533547e-05, - "loss": 2.4466, + "epoch": 0.10490274641218038, + "grad_norm": 0.96484375, + "learning_rate": 0.00019998674069827293, + "loss": 2.531, "step": 270 }, { - "epoch": 0.08678943154523619, - "grad_norm": 0.384765625, - "learning_rate": 4.3290734824281156e-05, - "loss": 2.4973, + "epoch": 0.10529127510259585, + "grad_norm": 0.8671875, + "learning_rate": 0.00019998443879587144, + "loss": 2.5153, "step": 271 }, { - "epoch": 0.08710968775020016, - "grad_norm": 0.5234375, - "learning_rate": 4.345047923322684e-05, - "loss": 2.3791, + "epoch": 0.10567980379301134, + "grad_norm": 0.87890625, + "learning_rate": 0.00019998195276111762, + "loss": 2.4719, "step": 272 }, { - "epoch": 0.08742994395516412, - "grad_norm": 0.455078125, - "learning_rate": 4.361022364217253e-05, - "loss": 2.2905, + "epoch": 0.10606833248342683, + "grad_norm": 0.859375, + "learning_rate": 0.00019997928259858985, + "loss": 2.4519, "step": 273 }, { - "epoch": 0.0877502001601281, - "grad_norm": 0.44140625, - "learning_rate": 4.376996805111822e-05, - "loss": 2.1515, + "epoch": 0.1064568611738423, + "grad_norm": 0.91796875, + "learning_rate": 0.00019997642831320547, + "loss": 2.3848, "step": 274 }, { - "epoch": 0.08807045636509207, - "grad_norm": 0.4921875, - "learning_rate": 4.39297124600639e-05, - "loss": 2.2641, + "epoch": 0.10684538986425779, + "grad_norm": 0.94921875, + "learning_rate": 0.00019997338991022096, + "loss": 2.4939, "step": 275 }, { - "epoch": 0.08839071257005604, - "grad_norm": 0.7265625, - "learning_rate": 4.408945686900959e-05, - "loss": 2.4807, + "epoch": 0.10723391855467328, + "grad_norm": 0.921875, + "learning_rate": 0.00019997016739523192, + "loss": 2.4016, "step": 276 }, { - "epoch": 0.08871096877502002, - "grad_norm": 0.419921875, - "learning_rate": 4.424920127795527e-05, - "loss": 2.3933, + "epoch": 0.10762244724508875, + "grad_norm": 0.9296875, + "learning_rate": 0.00019996676077417292, + "loss": 2.5356, "step": 277 }, { - "epoch": 0.08903122497998399, - "grad_norm": 0.53125, - "learning_rate": 4.440894568690096e-05, - "loss": 1.7232, + "epoch": 0.10801097593550424, + "grad_norm": 0.91796875, + "learning_rate": 0.00019996317005331767, + "loss": 2.5133, "step": 278 }, { - "epoch": 0.08935148118494796, - "grad_norm": 3.390625, - "learning_rate": 4.456869009584665e-05, - "loss": 2.3984, + "epoch": 0.10839950462591973, + "grad_norm": 0.96484375, + "learning_rate": 0.00019995939523927882, + "loss": 2.4938, "step": 279 }, { - "epoch": 0.08967173738991192, - "grad_norm": 0.4921875, - "learning_rate": 4.472843450479233e-05, - "loss": 2.2911, + "epoch": 0.1087880333163352, + "grad_norm": 0.87109375, + "learning_rate": 0.00019995543633900817, + "loss": 2.4887, "step": 280 }, { - "epoch": 0.0899919935948759, - "grad_norm": 0.5234375, - "learning_rate": 4.488817891373802e-05, - "loss": 2.3337, + "epoch": 0.10917656200675069, + "grad_norm": 0.96875, + "learning_rate": 0.00019995129335979644, + "loss": 2.4348, "step": 281 }, { - "epoch": 0.09031224979983987, - "grad_norm": 0.859375, - "learning_rate": 4.504792332268371e-05, - "loss": 2.0545, + "epoch": 0.10956509069716616, + "grad_norm": 0.91796875, + "learning_rate": 0.00019994696630927337, + "loss": 2.4854, "step": 282 }, { - "epoch": 0.09063250600480384, - "grad_norm": 0.52734375, - "learning_rate": 4.520766773162939e-05, - "loss": 2.1376, + "epoch": 0.10995361938758165, + "grad_norm": 1.1484375, + "learning_rate": 0.0001999424551954077, + "loss": 2.517, "step": 283 }, { - "epoch": 0.09095276220976782, - "grad_norm": 0.55859375, - "learning_rate": 4.536741214057508e-05, - "loss": 2.1938, + "epoch": 0.11034214807799714, + "grad_norm": 0.984375, + "learning_rate": 0.0001999377600265072, + "loss": 2.4481, "step": 284 }, { - "epoch": 0.09127301841473179, - "grad_norm": 0.53515625, - "learning_rate": 4.552715654952077e-05, - "loss": 2.4041, + "epoch": 0.11073067676841261, + "grad_norm": 1.046875, + "learning_rate": 0.00019993288081121845, + "loss": 2.4257, "step": 285 }, { - "epoch": 0.09159327461969576, - "grad_norm": 0.58203125, - "learning_rate": 4.568690095846646e-05, - "loss": 2.2543, + "epoch": 0.1111192054588281, + "grad_norm": 0.91796875, + "learning_rate": 0.00019992781755852712, + "loss": 2.5133, "step": 286 }, { - "epoch": 0.09191353082465972, - "grad_norm": 0.55859375, - "learning_rate": 4.584664536741215e-05, - "loss": 2.0115, + "epoch": 0.11150773414924359, + "grad_norm": 0.8828125, + "learning_rate": 0.00019992257027775777, + "loss": 2.5131, "step": 287 }, { - "epoch": 0.0922337870296237, - "grad_norm": 0.40234375, - "learning_rate": 4.600638977635783e-05, - "loss": 2.3996, + "epoch": 0.11189626283965906, + "grad_norm": 0.953125, + "learning_rate": 0.00019991713897857377, + "loss": 2.5679, "step": 288 }, { - "epoch": 0.09255404323458767, - "grad_norm": 2.546875, - "learning_rate": 4.616613418530352e-05, - "loss": 2.7077, + "epoch": 0.11228479153007455, + "grad_norm": 0.8984375, + "learning_rate": 0.00019991152367097753, + "loss": 2.4768, "step": 289 }, { - "epoch": 0.09287429943955164, - "grad_norm": 0.63671875, - "learning_rate": 4.632587859424921e-05, - "loss": 2.0203, + "epoch": 0.11267332022049004, + "grad_norm": 0.99609375, + "learning_rate": 0.0001999057243653102, + "loss": 2.4971, "step": 290 }, { - "epoch": 0.0931945556445156, - "grad_norm": 0.455078125, - "learning_rate": 4.648562300319489e-05, - "loss": 2.2094, + "epoch": 0.11306184891090551, + "grad_norm": 0.88671875, + "learning_rate": 0.00019989974107225184, + "loss": 2.4146, "step": 291 }, { - "epoch": 0.09351481184947959, - "grad_norm": 0.40234375, - "learning_rate": 4.664536741214058e-05, - "loss": 2.2682, + "epoch": 0.113450377601321, + "grad_norm": 0.90625, + "learning_rate": 0.0001998935738028214, + "loss": 2.4849, "step": 292 }, { - "epoch": 0.09383506805444355, - "grad_norm": 0.5234375, - "learning_rate": 4.680511182108626e-05, - "loss": 2.1374, + "epoch": 0.11383890629173649, + "grad_norm": 0.921875, + "learning_rate": 0.00019988722256837656, + "loss": 2.5275, "step": 293 }, { - "epoch": 0.09415532425940752, - "grad_norm": 0.9609375, - "learning_rate": 4.696485623003195e-05, - "loss": 2.272, + "epoch": 0.11422743498215196, + "grad_norm": 0.9765625, + "learning_rate": 0.0001998806873806138, + "loss": 2.4933, "step": 294 }, { - "epoch": 0.0944755804643715, - "grad_norm": 0.5078125, - "learning_rate": 4.712460063897764e-05, - "loss": 2.384, + "epoch": 0.11461596367256745, + "grad_norm": 0.8984375, + "learning_rate": 0.00019987396825156838, + "loss": 2.4731, "step": 295 }, { - "epoch": 0.09479583666933547, - "grad_norm": 0.671875, - "learning_rate": 4.728434504792332e-05, - "loss": 2.4198, + "epoch": 0.11500449236298294, + "grad_norm": 0.8984375, + "learning_rate": 0.00019986706519361434, + "loss": 2.489, "step": 296 }, { - "epoch": 0.09511609287429944, - "grad_norm": 0.3125, - "learning_rate": 4.744408945686901e-05, - "loss": 2.3007, + "epoch": 0.11539302105339841, + "grad_norm": 0.9453125, + "learning_rate": 0.00019985997821946443, + "loss": 2.4249, "step": 297 }, { - "epoch": 0.0954363490792634, - "grad_norm": 0.357421875, - "learning_rate": 4.76038338658147e-05, - "loss": 2.6156, + "epoch": 0.1157815497438139, + "grad_norm": 0.91015625, + "learning_rate": 0.00019985270734217006, + "loss": 2.4817, "step": 298 }, { - "epoch": 0.09575660528422739, - "grad_norm": 0.462890625, - "learning_rate": 4.776357827476038e-05, - "loss": 2.2731, + "epoch": 0.11617007843422937, + "grad_norm": 1.0546875, + "learning_rate": 0.00019984525257512142, + "loss": 2.5107, "step": 299 }, { - "epoch": 0.09607686148919135, - "grad_norm": 0.349609375, - "learning_rate": 4.792332268370607e-05, - "loss": 2.5316, + "epoch": 0.11655860712464486, + "grad_norm": 0.953125, + "learning_rate": 0.00019983761393204717, + "loss": 2.5681, "step": 300 }, { - "epoch": 0.09639711769415532, - "grad_norm": 3.75, - "learning_rate": 4.8083067092651754e-05, - "loss": 1.9556, + "epoch": 0.11694713581506035, + "grad_norm": 0.9609375, + "learning_rate": 0.0001998297914270148, + "loss": 2.5334, "step": 301 }, { - "epoch": 0.0967173738991193, - "grad_norm": 0.408203125, - "learning_rate": 4.824281150159744e-05, - "loss": 2.4008, + "epoch": 0.11733566450547582, + "grad_norm": 0.9453125, + "learning_rate": 0.00019982178507443028, + "loss": 2.4393, "step": 302 }, { - "epoch": 0.09703763010408327, - "grad_norm": 0.5703125, - "learning_rate": 4.840255591054313e-05, - "loss": 2.2126, + "epoch": 0.11772419319589131, + "grad_norm": 0.9140625, + "learning_rate": 0.0001998135948890382, + "loss": 2.4662, "step": 303 }, { - "epoch": 0.09735788630904724, - "grad_norm": 0.51171875, - "learning_rate": 4.856230031948882e-05, - "loss": 1.9912, + "epoch": 0.1181127218863068, + "grad_norm": 0.875, + "learning_rate": 0.0001998052208859217, + "loss": 2.4961, "step": 304 }, { - "epoch": 0.0976781425140112, - "grad_norm": 0.65625, - "learning_rate": 4.872204472843451e-05, - "loss": 2.3782, + "epoch": 0.11850125057672227, + "grad_norm": 0.90234375, + "learning_rate": 0.00019979666308050242, + "loss": 2.4746, "step": 305 }, { - "epoch": 0.09799839871897519, - "grad_norm": 0.94140625, - "learning_rate": 4.88817891373802e-05, - "loss": 2.2945, + "epoch": 0.11888977926713776, + "grad_norm": 1.015625, + "learning_rate": 0.0001997879214885405, + "loss": 2.508, "step": 306 }, { - "epoch": 0.09831865492393915, - "grad_norm": 0.416015625, - "learning_rate": 4.904153354632588e-05, - "loss": 2.7067, + "epoch": 0.11927830795755325, + "grad_norm": 0.83203125, + "learning_rate": 0.0001997789961261345, + "loss": 2.4174, "step": 307 }, { - "epoch": 0.09863891112890312, - "grad_norm": 0.423828125, - "learning_rate": 4.920127795527157e-05, - "loss": 2.0701, + "epoch": 0.11966683664796872, + "grad_norm": 1.0078125, + "learning_rate": 0.00019976988700972154, + "loss": 2.567, "step": 308 }, { - "epoch": 0.09895916733386709, - "grad_norm": 0.609375, - "learning_rate": 4.936102236421725e-05, - "loss": 2.5141, + "epoch": 0.1200553653383842, + "grad_norm": 0.875, + "learning_rate": 0.00019976059415607698, + "loss": 2.551, "step": 309 }, { - "epoch": 0.09927942353883107, - "grad_norm": 0.4609375, - "learning_rate": 4.952076677316294e-05, - "loss": 2.1579, + "epoch": 0.1204438940287997, + "grad_norm": 0.93359375, + "learning_rate": 0.0001997511175823147, + "loss": 2.5188, "step": 310 }, { - "epoch": 0.09959967974379504, - "grad_norm": 0.498046875, - "learning_rate": 4.968051118210863e-05, - "loss": 2.4784, + "epoch": 0.12083242271921517, + "grad_norm": 0.88671875, + "learning_rate": 0.00019974145730588684, + "loss": 2.5237, "step": 311 }, { - "epoch": 0.099919935948759, - "grad_norm": 0.498046875, - "learning_rate": 4.984025559105431e-05, - "loss": 2.285, + "epoch": 0.12122095140963066, + "grad_norm": 0.94140625, + "learning_rate": 0.00019973161334458386, + "loss": 2.4621, "step": 312 }, { - "epoch": 0.10024019215372298, - "grad_norm": 0.74609375, - "learning_rate": 5e-05, - "loss": 2.4196, + "epoch": 0.12160948010004614, + "grad_norm": 0.9296875, + "learning_rate": 0.00019972158571653444, + "loss": 2.4367, "step": 313 }, { - "epoch": 0.10056044835868695, - "grad_norm": 0.60546875, - "learning_rate": 4.9999984364699426e-05, - "loss": 1.4867, + "epoch": 0.12199800879046162, + "grad_norm": 0.890625, + "learning_rate": 0.00019971137444020563, + "loss": 2.3794, "step": 314 }, { - "epoch": 0.10088070456365092, - "grad_norm": 0.50390625, - "learning_rate": 4.999993745881724e-05, - "loss": 2.2844, + "epoch": 0.1223865374808771, + "grad_norm": 0.97265625, + "learning_rate": 0.00019970097953440262, + "loss": 2.402, "step": 315 }, { - "epoch": 0.10120096076861489, - "grad_norm": 0.703125, - "learning_rate": 4.9999859282412125e-05, - "loss": 1.8481, + "epoch": 0.12277506617129258, + "grad_norm": 1.03125, + "learning_rate": 0.00019969040101826873, + "loss": 2.4515, "step": 316 }, { - "epoch": 0.10152121697357887, - "grad_norm": 0.59375, - "learning_rate": 4.9999749835581864e-05, - "loss": 2.4302, + "epoch": 0.12316359486170807, + "grad_norm": 0.96484375, + "learning_rate": 0.00019967963891128547, + "loss": 2.4462, "step": 317 }, { - "epoch": 0.10184147317854284, - "grad_norm": 0.494140625, - "learning_rate": 4.999960911846336e-05, - "loss": 2.3678, + "epoch": 0.12355212355212356, + "grad_norm": 0.9453125, + "learning_rate": 0.00019966869323327248, + "loss": 2.4375, "step": 318 }, { - "epoch": 0.1021617293835068, - "grad_norm": 0.5546875, - "learning_rate": 4.999943713123261e-05, - "loss": 2.1427, + "epoch": 0.12394065224253903, + "grad_norm": 0.84375, + "learning_rate": 0.00019965756400438739, + "loss": 2.4505, "step": 319 }, { - "epoch": 0.10248198558847077, - "grad_norm": 0.47265625, - "learning_rate": 4.9999233874104755e-05, - "loss": 1.591, + "epoch": 0.12432918093295452, + "grad_norm": 1.1171875, + "learning_rate": 0.00019964625124512596, + "loss": 2.4783, "step": 320 }, { - "epoch": 0.10280224179343475, - "grad_norm": 0.4765625, - "learning_rate": 4.9998999347334036e-05, - "loss": 2.353, + "epoch": 0.12471770962337, + "grad_norm": 1.0234375, + "learning_rate": 0.00019963475497632182, + "loss": 2.4582, "step": 321 }, { - "epoch": 0.10312249799839872, - "grad_norm": 0.3984375, - "learning_rate": 4.9998733551213795e-05, - "loss": 2.0951, + "epoch": 0.12510623831378548, + "grad_norm": 0.91796875, + "learning_rate": 0.00019962307521914662, + "loss": 2.5833, "step": 322 }, { - "epoch": 0.10344275420336269, - "grad_norm": 0.5234375, - "learning_rate": 4.9998436486076495e-05, - "loss": 2.4225, + "epoch": 0.12549476700420098, + "grad_norm": 1.0703125, + "learning_rate": 0.00019961121199510998, + "loss": 2.4726, "step": 323 }, { - "epoch": 0.10376301040832667, - "grad_norm": 0.419921875, - "learning_rate": 4.999810815229372e-05, - "loss": 1.8435, + "epoch": 0.12588329569461645, + "grad_norm": 1.0234375, + "learning_rate": 0.00019959916532605925, + "loss": 2.5252, "step": 324 }, { - "epoch": 0.10408326661329063, - "grad_norm": 0.546875, - "learning_rate": 4.999774855027615e-05, - "loss": 2.0037, + "epoch": 0.12627182438503193, + "grad_norm": 0.95703125, + "learning_rate": 0.00019958693523417976, + "loss": 2.512, "step": 325 }, { - "epoch": 0.1044035228182546, - "grad_norm": 0.4453125, - "learning_rate": 4.999735768047359e-05, - "loss": 2.6103, + "epoch": 0.1266603530754474, + "grad_norm": 0.9765625, + "learning_rate": 0.00019957452174199452, + "loss": 2.4786, "step": 326 }, { - "epoch": 0.10472377902321857, - "grad_norm": 0.4765625, - "learning_rate": 4.999693554337495e-05, - "loss": 2.0573, + "epoch": 0.1270488817658629, + "grad_norm": 0.9453125, + "learning_rate": 0.0001995619248723644, + "loss": 2.5086, "step": 327 }, { - "epoch": 0.10504403522818255, - "grad_norm": 0.39453125, - "learning_rate": 4.9996482139508236e-05, - "loss": 2.1116, + "epoch": 0.12743741045627838, + "grad_norm": 0.91796875, + "learning_rate": 0.00019954914464848786, + "loss": 2.5146, "step": 328 }, { - "epoch": 0.10536429143314652, - "grad_norm": 0.37890625, - "learning_rate": 4.999599746944059e-05, - "loss": 2.677, + "epoch": 0.12782593914669385, + "grad_norm": 1.1484375, + "learning_rate": 0.00019953618109390115, + "loss": 2.4517, "step": 329 }, { - "epoch": 0.10568454763811048, - "grad_norm": 0.48046875, - "learning_rate": 4.9995481533778256e-05, - "loss": 2.2031, + "epoch": 0.12821446783710935, + "grad_norm": 0.94140625, + "learning_rate": 0.00019952303423247805, + "loss": 2.4473, "step": 330 }, { - "epoch": 0.10600480384307447, - "grad_norm": 0.6328125, - "learning_rate": 4.999493433316656e-05, - "loss": 2.1728, + "epoch": 0.12860299652752483, + "grad_norm": 1.1015625, + "learning_rate": 0.00019950970408842995, + "loss": 2.4946, "step": 331 }, { - "epoch": 0.10632506004803843, - "grad_norm": 0.494140625, - "learning_rate": 4.9994355868289965e-05, - "loss": 2.239, + "epoch": 0.1289915252179403, + "grad_norm": 0.9609375, + "learning_rate": 0.00019949619068630576, + "loss": 2.3974, "step": 332 }, { - "epoch": 0.1066453162530024, - "grad_norm": 0.431640625, - "learning_rate": 4.999374613987202e-05, - "loss": 2.5327, + "epoch": 0.1293800539083558, + "grad_norm": 1.0625, + "learning_rate": 0.00019948249405099197, + "loss": 2.5094, "step": 333 }, { - "epoch": 0.10696557245796637, - "grad_norm": 0.9765625, - "learning_rate": 4.9993105148675405e-05, - "loss": 2.2862, + "epoch": 0.12976858259877128, + "grad_norm": 1.0078125, + "learning_rate": 0.00019946861420771237, + "loss": 2.4857, "step": 334 }, { - "epoch": 0.10728582866293035, - "grad_norm": 0.36328125, - "learning_rate": 4.9992432895501874e-05, - "loss": 2.0678, + "epoch": 0.13015711128918675, + "grad_norm": 0.890625, + "learning_rate": 0.00019945455118202822, + "loss": 2.5822, "step": 335 }, { - "epoch": 0.10760608486789432, - "grad_norm": 0.671875, - "learning_rate": 4.99917293811923e-05, - "loss": 2.2222, + "epoch": 0.13054563997960225, + "grad_norm": 0.9296875, + "learning_rate": 0.0001994403049998382, + "loss": 2.4962, "step": 336 }, { - "epoch": 0.10792634107285828, - "grad_norm": 0.474609375, - "learning_rate": 4.9990994606626647e-05, - "loss": 2.195, + "epoch": 0.13093416867001773, + "grad_norm": 0.91015625, + "learning_rate": 0.00019942587568737817, + "loss": 2.5256, "step": 337 }, { - "epoch": 0.10824659727782225, - "grad_norm": 2.109375, - "learning_rate": 4.9990228572724e-05, - "loss": 2.0068, + "epoch": 0.1313226973604332, + "grad_norm": 0.97265625, + "learning_rate": 0.0001994112632712213, + "loss": 2.4939, "step": 338 }, { - "epoch": 0.10856685348278623, - "grad_norm": 0.88671875, - "learning_rate": 4.998943128044254e-05, - "loss": 2.0167, + "epoch": 0.1317112260508487, + "grad_norm": 0.859375, + "learning_rate": 0.00019939646777827793, + "loss": 2.4533, "step": 339 }, { - "epoch": 0.1088871096877502, - "grad_norm": 0.443359375, - "learning_rate": 4.9988602730779515e-05, - "loss": 2.3055, + "epoch": 0.13209975474126417, + "grad_norm": 0.8515625, + "learning_rate": 0.0001993814892357957, + "loss": 2.476, "step": 340 }, { - "epoch": 0.10920736589271417, - "grad_norm": 0.439453125, - "learning_rate": 4.998774292477131e-05, - "loss": 2.2542, + "epoch": 0.13248828343167965, + "grad_norm": 0.83203125, + "learning_rate": 0.0001993663276713591, + "loss": 2.4901, "step": 341 }, { - "epoch": 0.10952762209767815, - "grad_norm": 0.41015625, - "learning_rate": 4.99868518634934e-05, - "loss": 2.564, + "epoch": 0.13287681212209515, + "grad_norm": 0.85546875, + "learning_rate": 0.00019935098311288995, + "loss": 2.4806, "step": 342 }, { - "epoch": 0.10984787830264212, - "grad_norm": 0.4375, - "learning_rate": 4.998592954806033e-05, - "loss": 2.2191, + "epoch": 0.13326534081251062, + "grad_norm": 0.8984375, + "learning_rate": 0.00019933545558864686, + "loss": 2.4668, "step": 343 }, { - "epoch": 0.11016813450760608, - "grad_norm": 0.546875, - "learning_rate": 4.998497597962576e-05, - "loss": 2.5834, + "epoch": 0.1336538695029261, + "grad_norm": 0.84375, + "learning_rate": 0.0001993197451272255, + "loss": 2.5187, "step": 344 }, { - "epoch": 0.11048839071257005, - "grad_norm": 0.51171875, - "learning_rate": 4.998399115938244e-05, - "loss": 2.6079, + "epoch": 0.1340423981933416, + "grad_norm": 0.9140625, + "learning_rate": 0.0001993038517575584, + "loss": 2.5422, "step": 345 }, { - "epoch": 0.11080864691753403, - "grad_norm": 0.6171875, - "learning_rate": 4.99829750885622e-05, - "loss": 2.0724, + "epoch": 0.13443092688375707, + "grad_norm": 0.9375, + "learning_rate": 0.000199287775508915, + "loss": 2.472, "step": 346 }, { - "epoch": 0.111128903122498, - "grad_norm": 0.51171875, - "learning_rate": 4.9981927768435964e-05, - "loss": 2.0157, + "epoch": 0.13481945557417255, + "grad_norm": 0.9296875, + "learning_rate": 0.00019927151641090145, + "loss": 2.4643, "step": 347 }, { - "epoch": 0.11144915932746197, - "grad_norm": 0.58203125, - "learning_rate": 4.998084920031376e-05, - "loss": 2.4644, + "epoch": 0.13520798426458805, + "grad_norm": 0.92578125, + "learning_rate": 0.00019925507449346065, + "loss": 2.4486, "step": 348 }, { - "epoch": 0.11176941553242595, - "grad_norm": 0.388671875, - "learning_rate": 4.997973938554466e-05, - "loss": 2.6107, + "epoch": 0.13559651295500352, + "grad_norm": 0.94921875, + "learning_rate": 0.00019923844978687224, + "loss": 2.4472, "step": 349 }, { - "epoch": 0.11208967173738991, - "grad_norm": 0.4921875, - "learning_rate": 4.997859832551689e-05, - "loss": 2.3877, + "epoch": 0.135985041645419, + "grad_norm": 0.95703125, + "learning_rate": 0.0001992216423217524, + "loss": 2.3693, "step": 350 }, { - "epoch": 0.11240992794235388, - "grad_norm": 0.466796875, - "learning_rate": 4.997742602165767e-05, - "loss": 2.179, + "epoch": 0.1363735703358345, + "grad_norm": 0.9921875, + "learning_rate": 0.00019920465212905394, + "loss": 2.4148, "step": 351 }, { - "epoch": 0.11273018414731785, - "grad_norm": 1.2890625, - "learning_rate": 4.997622247543338e-05, - "loss": 2.5496, + "epoch": 0.13676209902624997, + "grad_norm": 1.046875, + "learning_rate": 0.00019918747924006623, + "loss": 2.559, "step": 352 }, { - "epoch": 0.11305044035228183, - "grad_norm": 0.515625, - "learning_rate": 4.9974987688349415e-05, - "loss": 1.8469, + "epoch": 0.13715062771666545, + "grad_norm": 1.0546875, + "learning_rate": 0.00019917012368641496, + "loss": 2.433, "step": 353 }, { - "epoch": 0.1133706965572458, - "grad_norm": 0.5703125, - "learning_rate": 4.9973721661950295e-05, - "loss": 2.3798, + "epoch": 0.13753915640708095, + "grad_norm": 1.234375, + "learning_rate": 0.00019915258550006235, + "loss": 2.4333, "step": 354 }, { - "epoch": 0.11369095276220977, - "grad_norm": 0.875, - "learning_rate": 4.9972424397819596e-05, - "loss": 2.4496, + "epoch": 0.13792768509749642, + "grad_norm": 0.93359375, + "learning_rate": 0.00019913486471330685, + "loss": 2.4635, "step": 355 }, { - "epoch": 0.11401120896717373, - "grad_norm": 0.47265625, - "learning_rate": 4.997109589757997e-05, - "loss": 2.0667, + "epoch": 0.1383162137879119, + "grad_norm": 0.97265625, + "learning_rate": 0.0001991169613587833, + "loss": 2.4907, "step": 356 }, { - "epoch": 0.11433146517213771, - "grad_norm": 3.4375, - "learning_rate": 4.996973616289312e-05, - "loss": 2.1528, + "epoch": 0.1387047424783274, + "grad_norm": 1.0, + "learning_rate": 0.00019909887546946261, + "loss": 2.4457, "step": 357 }, { - "epoch": 0.11465172137710168, - "grad_norm": 0.4609375, - "learning_rate": 4.996834519545985e-05, - "loss": 1.8263, + "epoch": 0.13909327116874287, + "grad_norm": 0.86328125, + "learning_rate": 0.00019908060707865202, + "loss": 2.4854, "step": 358 }, { - "epoch": 0.11497197758206565, - "grad_norm": 0.466796875, - "learning_rate": 4.996692299702e-05, - "loss": 2.629, + "epoch": 0.13948179985915835, + "grad_norm": 0.9140625, + "learning_rate": 0.00019906215621999474, + "loss": 2.4735, "step": 359 }, { - "epoch": 0.11529223378702963, - "grad_norm": 0.55859375, - "learning_rate": 4.996546956935252e-05, - "loss": 2.4788, + "epoch": 0.13987032854957382, + "grad_norm": 0.9140625, + "learning_rate": 0.00019904352292747008, + "loss": 2.537, "step": 360 }, { - "epoch": 0.1156124899919936, - "grad_norm": 0.41015625, - "learning_rate": 4.9963984914275363e-05, - "loss": 2.2397, + "epoch": 0.14025885723998932, + "grad_norm": 0.88671875, + "learning_rate": 0.00019902470723539328, + "loss": 2.5154, "step": 361 }, { - "epoch": 0.11593274619695756, - "grad_norm": 0.5078125, - "learning_rate": 4.996246903364559e-05, - "loss": 2.3035, + "epoch": 0.1406473859304048, + "grad_norm": 0.96875, + "learning_rate": 0.00019900570917841552, + "loss": 2.4888, "step": 362 }, { - "epoch": 0.11625300240192153, - "grad_norm": 0.64453125, - "learning_rate": 4.996092192935929e-05, - "loss": 2.6009, + "epoch": 0.14103591462082027, + "grad_norm": 0.88671875, + "learning_rate": 0.0001989865287915238, + "loss": 2.4895, "step": 363 }, { - "epoch": 0.11657325860688551, - "grad_norm": 0.52734375, - "learning_rate": 4.99593436033516e-05, - "loss": 2.3057, + "epoch": 0.14142444331123577, + "grad_norm": 0.86328125, + "learning_rate": 0.00019896716611004096, + "loss": 2.464, "step": 364 }, { - "epoch": 0.11689351481184948, - "grad_norm": 0.42578125, - "learning_rate": 4.9957734057596774e-05, - "loss": 2.5767, + "epoch": 0.14181297200165124, + "grad_norm": 0.8828125, + "learning_rate": 0.00019894762116962544, + "loss": 2.4514, "step": 365 }, { - "epoch": 0.11721377101681345, - "grad_norm": 0.466796875, - "learning_rate": 4.995609329410804e-05, - "loss": 1.9374, + "epoch": 0.14220150069206672, + "grad_norm": 1.03125, + "learning_rate": 0.00019892789400627147, + "loss": 2.455, "step": 366 }, { - "epoch": 0.11753402722177742, - "grad_norm": 0.640625, - "learning_rate": 4.995442131493771e-05, - "loss": 1.9711, + "epoch": 0.14259002938248222, + "grad_norm": 0.8671875, + "learning_rate": 0.0001989079846563087, + "loss": 2.4248, "step": 367 }, { - "epoch": 0.1178542834267414, - "grad_norm": 0.70703125, - "learning_rate": 4.9952718122177126e-05, - "loss": 2.3285, + "epoch": 0.1429785580728977, + "grad_norm": 0.953125, + "learning_rate": 0.00019888789315640254, + "loss": 2.5139, "step": 368 }, { - "epoch": 0.11817453963170536, - "grad_norm": 0.546875, - "learning_rate": 4.9950983717956703e-05, - "loss": 2.0349, + "epoch": 0.14336708676331317, + "grad_norm": 0.84375, + "learning_rate": 0.00019886761954355354, + "loss": 2.4619, "step": 369 }, { - "epoch": 0.11849479583666933, - "grad_norm": 0.578125, - "learning_rate": 4.994921810444586e-05, - "loss": 2.1336, + "epoch": 0.14375561545372867, + "grad_norm": 0.90234375, + "learning_rate": 0.00019884716385509785, + "loss": 2.517, "step": 370 }, { - "epoch": 0.11881505204163331, - "grad_norm": 0.447265625, - "learning_rate": 4.994742128385307e-05, - "loss": 2.319, + "epoch": 0.14414414414414414, + "grad_norm": 0.9375, + "learning_rate": 0.0001988265261287069, + "loss": 2.474, "step": 371 }, { - "epoch": 0.11913530824659728, - "grad_norm": 0.408203125, - "learning_rate": 4.994559325842585e-05, - "loss": 2.0951, + "epoch": 0.14453267283455962, + "grad_norm": 0.84375, + "learning_rate": 0.00019880570640238733, + "loss": 2.511, "step": 372 }, { - "epoch": 0.11945556445156125, - "grad_norm": 0.9765625, - "learning_rate": 4.994373403045073e-05, - "loss": 2.4311, + "epoch": 0.14492120152497512, + "grad_norm": 0.89453125, + "learning_rate": 0.00019878470471448094, + "loss": 2.4679, "step": 373 }, { - "epoch": 0.11977582065652521, - "grad_norm": 0.6484375, - "learning_rate": 4.994184360225328e-05, - "loss": 1.8299, + "epoch": 0.1453097302153906, + "grad_norm": 0.85546875, + "learning_rate": 0.00019876352110366466, + "loss": 2.4047, "step": 374 }, { - "epoch": 0.1200960768614892, - "grad_norm": 0.435546875, - "learning_rate": 4.993992197619809e-05, - "loss": 2.4937, + "epoch": 0.14569825890580607, + "grad_norm": 0.93359375, + "learning_rate": 0.0001987421556089504, + "loss": 2.5233, "step": 375 }, { - "epoch": 0.12041633306645316, - "grad_norm": 0.486328125, - "learning_rate": 4.9937969154688766e-05, - "loss": 2.3098, + "epoch": 0.14608678759622157, + "grad_norm": 0.88671875, + "learning_rate": 0.00019872060826968513, + "loss": 2.4157, "step": 376 }, { - "epoch": 0.12073658927141713, - "grad_norm": 0.55078125, - "learning_rate": 4.993598514016797e-05, - "loss": 2.2064, + "epoch": 0.14647531628663704, + "grad_norm": 0.94140625, + "learning_rate": 0.00019869887912555062, + "loss": 2.4568, "step": 377 }, { - "epoch": 0.12105684547638111, - "grad_norm": 0.48828125, - "learning_rate": 4.9933969935117335e-05, - "loss": 2.0565, + "epoch": 0.14686384497705252, + "grad_norm": 0.87890625, + "learning_rate": 0.0001986769682165635, + "loss": 2.4597, "step": 378 }, { - "epoch": 0.12137710168134508, - "grad_norm": 0.453125, - "learning_rate": 4.9931923542057534e-05, - "loss": 2.6077, + "epoch": 0.14725237366746802, + "grad_norm": 0.890625, + "learning_rate": 0.0001986548755830751, + "loss": 2.5093, "step": 379 }, { - "epoch": 0.12169735788630905, - "grad_norm": 0.375, - "learning_rate": 4.9929845963548246e-05, - "loss": 2.3129, + "epoch": 0.1476409023578835, + "grad_norm": 0.89453125, + "learning_rate": 0.00019863260126577146, + "loss": 2.4912, "step": 380 }, { - "epoch": 0.12201761409127301, - "grad_norm": 0.48046875, - "learning_rate": 4.9927737202188155e-05, - "loss": 2.5353, + "epoch": 0.14802943104829897, + "grad_norm": 0.875, + "learning_rate": 0.00019861014530567314, + "loss": 2.5111, "step": 381 }, { - "epoch": 0.122337870296237, - "grad_norm": 0.439453125, - "learning_rate": 4.992559726061494e-05, - "loss": 2.0995, + "epoch": 0.14841795973871447, + "grad_norm": 0.83203125, + "learning_rate": 0.0001985875077441353, + "loss": 2.5189, "step": 382 }, { - "epoch": 0.12265812650120096, - "grad_norm": 0.41796875, - "learning_rate": 4.9923426141505305e-05, - "loss": 1.9616, + "epoch": 0.14880648842912994, + "grad_norm": 0.85546875, + "learning_rate": 0.00019856468862284752, + "loss": 2.4845, "step": 383 }, { - "epoch": 0.12297838270616493, - "grad_norm": 0.40625, - "learning_rate": 4.9921223847574935e-05, - "loss": 2.1245, + "epoch": 0.14919501711954541, + "grad_norm": 0.88671875, + "learning_rate": 0.0001985416879838337, + "loss": 2.5138, "step": 384 }, { - "epoch": 0.1232986389111289, - "grad_norm": 0.7734375, - "learning_rate": 4.991899038157851e-05, - "loss": 2.2955, + "epoch": 0.14958354580996092, + "grad_norm": 0.9140625, + "learning_rate": 0.00019851850586945203, + "loss": 2.459, "step": 385 }, { - "epoch": 0.12361889511609288, - "grad_norm": 0.59765625, - "learning_rate": 4.991672574630971e-05, - "loss": 2.2364, + "epoch": 0.1499720745003764, + "grad_norm": 0.890625, + "learning_rate": 0.00019849514232239495, + "loss": 2.4799, "step": 386 }, { - "epoch": 0.12393915132105685, - "grad_norm": 0.4609375, - "learning_rate": 4.9914429944601185e-05, - "loss": 2.2836, + "epoch": 0.15036060319079186, + "grad_norm": 0.921875, + "learning_rate": 0.000198471597385689, + "loss": 2.4408, "step": 387 }, { - "epoch": 0.12425940752602081, - "grad_norm": 0.486328125, - "learning_rate": 4.991210297932459e-05, - "loss": 1.9992, + "epoch": 0.15074913188120737, + "grad_norm": 0.8046875, + "learning_rate": 0.00019844787110269478, + "loss": 2.4064, "step": 388 }, { - "epoch": 0.1245796637309848, - "grad_norm": 0.33984375, - "learning_rate": 4.9909744853390534e-05, - "loss": 2.2879, + "epoch": 0.15113766057162284, + "grad_norm": 0.95703125, + "learning_rate": 0.00019842396351710685, + "loss": 2.4311, "step": 389 }, { - "epoch": 0.12489991993594876, - "grad_norm": 0.87109375, - "learning_rate": 4.990735556974863e-05, - "loss": 2.2375, + "epoch": 0.1515261892620383, + "grad_norm": 0.83984375, + "learning_rate": 0.00019839987467295362, + "loss": 2.4982, "step": 390 }, { - "epoch": 0.12522017614091274, - "grad_norm": 0.59375, - "learning_rate": 4.990493513138744e-05, - "loss": 2.5697, + "epoch": 0.15191471795245381, + "grad_norm": 0.83203125, + "learning_rate": 0.00019837560461459744, + "loss": 2.504, "step": 391 }, { - "epoch": 0.1255404323458767, - "grad_norm": 0.40234375, - "learning_rate": 4.990248354133452e-05, - "loss": 1.9927, + "epoch": 0.1523032466428693, + "grad_norm": 0.84765625, + "learning_rate": 0.0001983511533867342, + "loss": 2.5254, "step": 392 }, { - "epoch": 0.12586068855084068, - "grad_norm": 0.5, - "learning_rate": 4.9900000802656376e-05, - "loss": 2.2876, + "epoch": 0.15269177533328476, + "grad_norm": 1.0078125, + "learning_rate": 0.0001983265210343936, + "loss": 2.4947, "step": 393 }, { - "epoch": 0.12618094475580463, - "grad_norm": 0.53515625, - "learning_rate": 4.989748691845847e-05, - "loss": 2.3259, + "epoch": 0.15308030402370024, + "grad_norm": 0.921875, + "learning_rate": 0.00019830170760293877, + "loss": 2.4509, "step": 394 }, { - "epoch": 0.1265012009607686, - "grad_norm": 0.53515625, - "learning_rate": 4.989494189188523e-05, - "loss": 2.6403, + "epoch": 0.15346883271411574, + "grad_norm": 0.90625, + "learning_rate": 0.00019827671313806642, + "loss": 2.4467, "step": 395 }, { - "epoch": 0.1268214571657326, - "grad_norm": 0.84765625, - "learning_rate": 4.989236572612004e-05, - "loss": 2.4835, + "epoch": 0.1538573614045312, + "grad_norm": 0.79296875, + "learning_rate": 0.0001982515376858066, + "loss": 2.4584, "step": 396 }, { - "epoch": 0.12714171337069655, - "grad_norm": 0.45703125, - "learning_rate": 4.988975842438523e-05, - "loss": 2.2297, + "epoch": 0.15424589009494669, + "grad_norm": 0.94140625, + "learning_rate": 0.00019822618129252263, + "loss": 2.4464, "step": 397 }, { - "epoch": 0.12746196957566053, - "grad_norm": 0.65234375, - "learning_rate": 4.9887119989942073e-05, - "loss": 2.4837, + "epoch": 0.1546344187853622, + "grad_norm": 0.8515625, + "learning_rate": 0.00019820064400491116, + "loss": 2.471, "step": 398 }, { - "epoch": 0.1277822257806245, - "grad_norm": 0.478515625, - "learning_rate": 4.9884450426090786e-05, - "loss": 2.9118, + "epoch": 0.15502294747577766, + "grad_norm": 0.90234375, + "learning_rate": 0.0001981749258700019, + "loss": 2.4903, "step": 399 }, { - "epoch": 0.12810248198558846, - "grad_norm": 0.51171875, - "learning_rate": 4.9881749736170525e-05, - "loss": 2.4227, + "epoch": 0.15541147616619314, + "grad_norm": 0.85546875, + "learning_rate": 0.00019814902693515764, + "loss": 2.4695, "step": 400 }, { - "epoch": 0.12842273819055244, - "grad_norm": 0.5234375, - "learning_rate": 4.987901792355938e-05, - "loss": 2.6651, + "epoch": 0.15580000485660864, + "grad_norm": 0.83984375, + "learning_rate": 0.00019812294724807407, + "loss": 2.5074, "step": 401 }, { - "epoch": 0.12874299439551642, - "grad_norm": 0.55859375, - "learning_rate": 4.987625499167436e-05, - "loss": 2.61, + "epoch": 0.1561885335470241, + "grad_norm": 0.89453125, + "learning_rate": 0.00019809668685677988, + "loss": 2.5358, "step": 402 }, { - "epoch": 0.12906325060048038, - "grad_norm": 0.341796875, - "learning_rate": 4.9873460943971416e-05, - "loss": 2.1197, + "epoch": 0.15657706223743958, + "grad_norm": 0.87109375, + "learning_rate": 0.0001980702458096364, + "loss": 2.5308, "step": 403 }, { - "epoch": 0.12938350680544436, - "grad_norm": 0.58984375, - "learning_rate": 4.987063578394541e-05, - "loss": 2.6976, + "epoch": 0.1569655909278551, + "grad_norm": 0.84375, + "learning_rate": 0.00019804362415533775, + "loss": 2.4968, "step": 404 }, { - "epoch": 0.1297037630104083, - "grad_norm": 0.65234375, - "learning_rate": 4.986777951513011e-05, - "loss": 2.242, + "epoch": 0.15735411961827056, + "grad_norm": 0.81640625, + "learning_rate": 0.00019801682194291068, + "loss": 2.478, "step": 405 }, { - "epoch": 0.1300240192153723, - "grad_norm": 0.46484375, - "learning_rate": 4.9864892141098215e-05, - "loss": 2.3003, + "epoch": 0.15774264830868603, + "grad_norm": 0.9140625, + "learning_rate": 0.00019798983922171437, + "loss": 2.4332, "step": 406 }, { - "epoch": 0.13034427542033628, - "grad_norm": 0.396484375, - "learning_rate": 4.986197366546133e-05, - "loss": 1.8946, + "epoch": 0.15813117699910154, + "grad_norm": 0.99609375, + "learning_rate": 0.00019796267604144045, + "loss": 2.5514, "step": 407 }, { - "epoch": 0.13066453162530023, - "grad_norm": 0.349609375, - "learning_rate": 4.985902409186993e-05, - "loss": 2.1781, + "epoch": 0.158519705689517, + "grad_norm": 0.859375, + "learning_rate": 0.00019793533245211298, + "loss": 2.4069, "step": 408 }, { - "epoch": 0.1309847878302642, - "grad_norm": 0.515625, - "learning_rate": 4.985604342401343e-05, - "loss": 1.9627, + "epoch": 0.15890823437993248, + "grad_norm": 0.8671875, + "learning_rate": 0.00019790780850408816, + "loss": 2.508, "step": 409 }, { - "epoch": 0.1313050440352282, - "grad_norm": 0.4140625, - "learning_rate": 4.9853031665620116e-05, - "loss": 2.161, + "epoch": 0.15929676307034799, + "grad_norm": 0.86328125, + "learning_rate": 0.00019788010424805433, + "loss": 2.4841, "step": 410 }, { - "epoch": 0.13162530024019214, - "grad_norm": 1.7734375, - "learning_rate": 4.984998882045717e-05, - "loss": 2.3552, + "epoch": 0.15968529176076346, + "grad_norm": 0.88671875, + "learning_rate": 0.00019785221973503197, + "loss": 2.4688, "step": 411 }, { - "epoch": 0.13194555644515613, - "grad_norm": 0.439453125, - "learning_rate": 4.9846914892330654e-05, - "loss": 2.2441, + "epoch": 0.16007382045117893, + "grad_norm": 0.90625, + "learning_rate": 0.00019782415501637347, + "loss": 2.4655, "step": 412 }, { - "epoch": 0.1322658126501201, - "grad_norm": 0.396484375, - "learning_rate": 4.984380988508551e-05, - "loss": 1.9077, + "epoch": 0.16046234914159443, + "grad_norm": 0.93359375, + "learning_rate": 0.0001977959101437631, + "loss": 2.5057, "step": 413 }, { - "epoch": 0.13258606885508406, - "grad_norm": 0.48828125, - "learning_rate": 4.9840673802605566e-05, - "loss": 2.5129, + "epoch": 0.1608508778320099, + "grad_norm": 0.890625, + "learning_rate": 0.0001977674851692169, + "loss": 2.4822, "step": 414 }, { - "epoch": 0.13290632506004804, - "grad_norm": 0.46875, - "learning_rate": 4.98375066488135e-05, - "loss": 2.2193, + "epoch": 0.16123940652242538, + "grad_norm": 0.94140625, + "learning_rate": 0.00019773888014508259, + "loss": 2.4492, "step": 415 }, { - "epoch": 0.13322658126501202, - "grad_norm": 0.48828125, - "learning_rate": 4.983430842767086e-05, - "loss": 2.5459, + "epoch": 0.16162793521284088, + "grad_norm": 0.91796875, + "learning_rate": 0.00019771009512403948, + "loss": 2.4674, "step": 416 }, { - "epoch": 0.13354683746997598, - "grad_norm": 0.490234375, - "learning_rate": 4.9831079143178066e-05, - "loss": 2.6023, + "epoch": 0.16201646390325636, + "grad_norm": 0.9453125, + "learning_rate": 0.00019768113015909835, + "loss": 2.489, "step": 417 }, { - "epoch": 0.13386709367493996, - "grad_norm": 9.5, - "learning_rate": 4.982781879937438e-05, - "loss": 2.1457, + "epoch": 0.16240499259367183, + "grad_norm": 0.859375, + "learning_rate": 0.00019765198530360137, + "loss": 2.4627, "step": 418 }, { - "epoch": 0.1341873498799039, - "grad_norm": 0.404296875, - "learning_rate": 4.982452740033793e-05, - "loss": 2.0819, + "epoch": 0.16279352128408733, + "grad_norm": 0.84765625, + "learning_rate": 0.000197622660611222, + "loss": 2.4688, "step": 419 }, { - "epoch": 0.1345076060848679, - "grad_norm": 0.703125, - "learning_rate": 4.982120495018566e-05, - "loss": 2.6695, + "epoch": 0.1631820499745028, + "grad_norm": 0.859375, + "learning_rate": 0.00019759315613596492, + "loss": 2.4542, "step": 420 }, { - "epoch": 0.13482786228983187, - "grad_norm": 0.59765625, - "learning_rate": 4.981785145307337e-05, - "loss": 2.4198, + "epoch": 0.16357057866491828, + "grad_norm": 0.87890625, + "learning_rate": 0.0001975634719321658, + "loss": 2.4351, "step": 421 }, { - "epoch": 0.13514811849479583, - "grad_norm": 0.6171875, - "learning_rate": 4.9814466913195717e-05, - "loss": 2.0445, + "epoch": 0.16395910735533378, + "grad_norm": 0.8828125, + "learning_rate": 0.00019753360805449145, + "loss": 2.4476, "step": 422 }, { - "epoch": 0.1354683746997598, - "grad_norm": 0.40625, - "learning_rate": 4.981105133478614e-05, - "loss": 2.3541, + "epoch": 0.16434763604574926, + "grad_norm": 0.890625, + "learning_rate": 0.0001975035645579394, + "loss": 2.4848, "step": 423 }, { - "epoch": 0.1357886309047238, - "grad_norm": 0.34375, - "learning_rate": 4.9807604722116945e-05, - "loss": 1.7954, + "epoch": 0.16473616473616473, + "grad_norm": 0.94140625, + "learning_rate": 0.00019747334149783813, + "loss": 2.5401, "step": 424 }, { - "epoch": 0.13610888710968774, - "grad_norm": 0.640625, - "learning_rate": 4.980412707949923e-05, - "loss": 2.3201, + "epoch": 0.16512469342658023, + "grad_norm": 0.8515625, + "learning_rate": 0.0001974429389298467, + "loss": 2.4539, "step": 425 }, { - "epoch": 0.13642914331465172, - "grad_norm": 0.67578125, - "learning_rate": 4.9800618411282914e-05, - "loss": 2.3104, + "epoch": 0.1655132221169957, + "grad_norm": 0.8984375, + "learning_rate": 0.0001974123569099548, + "loss": 2.4487, "step": 426 }, { - "epoch": 0.1367493995196157, - "grad_norm": 0.41796875, - "learning_rate": 4.979707872185672e-05, - "loss": 2.4099, + "epoch": 0.16590175080741118, + "grad_norm": 0.83984375, + "learning_rate": 0.0001973815954944826, + "loss": 2.4444, "step": 427 }, { - "epoch": 0.13706965572457966, - "grad_norm": 0.3671875, - "learning_rate": 4.979350801564818e-05, - "loss": 2.0, + "epoch": 0.16629027949782665, + "grad_norm": 1.0078125, + "learning_rate": 0.00019735065474008062, + "loss": 2.4784, "step": 428 }, { - "epoch": 0.13738991192954364, - "grad_norm": 0.466796875, - "learning_rate": 4.9789906297123626e-05, - "loss": 2.3145, + "epoch": 0.16667880818824216, + "grad_norm": 0.91796875, + "learning_rate": 0.00019731953470372964, + "loss": 2.4598, "step": 429 }, { - "epoch": 0.1377101681345076, - "grad_norm": 0.451171875, - "learning_rate": 4.978627357078817e-05, - "loss": 2.1489, + "epoch": 0.16706733687865763, + "grad_norm": 0.97265625, + "learning_rate": 0.00019728823544274063, + "loss": 2.4388, "step": 430 }, { - "epoch": 0.13803042433947157, - "grad_norm": 0.474609375, - "learning_rate": 4.97826098411857e-05, - "loss": 2.0548, + "epoch": 0.1674558655690731, + "grad_norm": 0.95703125, + "learning_rate": 0.00019725675701475468, + "loss": 2.4653, "step": 431 }, { - "epoch": 0.13835068054443556, - "grad_norm": 0.7890625, - "learning_rate": 4.9778915112898914e-05, - "loss": 2.3199, + "epoch": 0.1678443942594886, + "grad_norm": 0.875, + "learning_rate": 0.00019722509947774276, + "loss": 2.4829, "step": 432 }, { - "epoch": 0.1386709367493995, - "grad_norm": 12.4375, - "learning_rate": 4.977518939054927e-05, - "loss": 2.1056, + "epoch": 0.16823292294990408, + "grad_norm": 0.875, + "learning_rate": 0.00019719326289000568, + "loss": 2.4322, "step": 433 }, { - "epoch": 0.1389911929543635, - "grad_norm": 0.443359375, - "learning_rate": 4.977143267879697e-05, - "loss": 2.2003, + "epoch": 0.16862145164031955, + "grad_norm": 1.0546875, + "learning_rate": 0.00019716124731017405, + "loss": 2.4855, "step": 434 }, { - "epoch": 0.13931144915932747, - "grad_norm": 0.51953125, - "learning_rate": 4.9767644982341014e-05, - "loss": 2.1945, + "epoch": 0.16900998033073505, + "grad_norm": 0.8828125, + "learning_rate": 0.00019712905279720806, + "loss": 2.4812, "step": 435 }, { - "epoch": 0.13963170536429143, - "grad_norm": 0.47265625, - "learning_rate": 4.9763826305919146e-05, - "loss": 2.4707, + "epoch": 0.16939850902115053, + "grad_norm": 0.859375, + "learning_rate": 0.0001970966794103975, + "loss": 2.4207, "step": 436 }, { - "epoch": 0.1399519615692554, - "grad_norm": 0.455078125, - "learning_rate": 4.975997665430785e-05, - "loss": 2.5856, + "epoch": 0.169787037711566, + "grad_norm": 0.90234375, + "learning_rate": 0.00019706412720936145, + "loss": 2.3899, "step": 437 }, { - "epoch": 0.1402722177742194, - "grad_norm": 0.4921875, - "learning_rate": 4.9756096032322376e-05, - "loss": 2.2698, + "epoch": 0.1701755664019815, + "grad_norm": 0.890625, + "learning_rate": 0.00019703139625404844, + "loss": 2.4568, "step": 438 }, { - "epoch": 0.14059247397918334, - "grad_norm": 0.34765625, - "learning_rate": 4.975218444481668e-05, - "loss": 2.5266, + "epoch": 0.17056409509239698, + "grad_norm": 0.90234375, + "learning_rate": 0.00019699848660473612, + "loss": 2.495, "step": 439 }, { - "epoch": 0.14091273018414732, - "grad_norm": 0.5859375, - "learning_rate": 4.974824189668348e-05, - "loss": 2.2826, + "epoch": 0.17095262378281245, + "grad_norm": 0.85546875, + "learning_rate": 0.0001969653983220312, + "loss": 2.5276, "step": 440 }, { - "epoch": 0.14123298638911128, - "grad_norm": 1.28125, - "learning_rate": 4.974426839285422e-05, - "loss": 1.8341, + "epoch": 0.17134115247322795, + "grad_norm": 0.87109375, + "learning_rate": 0.0001969321314668694, + "loss": 2.4646, "step": 441 }, { - "epoch": 0.14155324259407526, - "grad_norm": 0.439453125, - "learning_rate": 4.9740263938299034e-05, - "loss": 2.6753, + "epoch": 0.17172968116364343, + "grad_norm": 0.8984375, + "learning_rate": 0.0001968986861005153, + "loss": 2.4508, "step": 442 }, { - "epoch": 0.14187349879903924, - "grad_norm": 0.458984375, - "learning_rate": 4.973622853802681e-05, - "loss": 2.436, + "epoch": 0.1721182098540589, + "grad_norm": 0.87890625, + "learning_rate": 0.00019686506228456225, + "loss": 2.5221, "step": 443 }, { - "epoch": 0.1421937550040032, - "grad_norm": 0.439453125, - "learning_rate": 4.97321621970851e-05, - "loss": 2.1716, + "epoch": 0.1725067385444744, + "grad_norm": 0.859375, + "learning_rate": 0.0001968312600809322, + "loss": 2.4256, "step": 444 }, { - "epoch": 0.14251401120896717, - "grad_norm": 0.62109375, - "learning_rate": 4.97280649205602e-05, - "loss": 2.0777, + "epoch": 0.17289526723488988, + "grad_norm": 0.828125, + "learning_rate": 0.0001967972795518756, + "loss": 2.4289, "step": 445 }, { - "epoch": 0.14283426741393115, - "grad_norm": 0.41015625, - "learning_rate": 4.9723936713577084e-05, - "loss": 1.7713, + "epoch": 0.17328379592530535, + "grad_norm": 0.91796875, + "learning_rate": 0.00019676312075997139, + "loss": 2.374, "step": 446 }, { - "epoch": 0.1431545236188951, - "grad_norm": 0.52734375, - "learning_rate": 4.97197775812994e-05, - "loss": 2.645, + "epoch": 0.17367232461572085, + "grad_norm": 0.8515625, + "learning_rate": 0.00019672878376812667, + "loss": 2.4398, "step": 447 }, { - "epoch": 0.1434747798238591, - "grad_norm": 0.53515625, - "learning_rate": 4.97155875289295e-05, - "loss": 2.1758, + "epoch": 0.17406085330613633, + "grad_norm": 0.8828125, + "learning_rate": 0.00019669426863957686, + "loss": 2.3595, "step": 448 }, { - "epoch": 0.14379503602882307, - "grad_norm": 0.447265625, - "learning_rate": 4.9711366561708395e-05, - "loss": 2.2733, + "epoch": 0.1744493819965518, + "grad_norm": 0.8515625, + "learning_rate": 0.00019665957543788532, + "loss": 2.4631, "step": 449 }, { - "epoch": 0.14411529223378702, - "grad_norm": 0.42578125, - "learning_rate": 4.9707114684915776e-05, - "loss": 1.9709, + "epoch": 0.1748379106869673, + "grad_norm": 0.83984375, + "learning_rate": 0.00019662470422694342, + "loss": 2.4148, "step": 450 }, { - "epoch": 0.144435548438751, - "grad_norm": 0.279296875, - "learning_rate": 4.9702831903869996e-05, - "loss": 1.9222, + "epoch": 0.17522643937738278, + "grad_norm": 0.97265625, + "learning_rate": 0.0001965896550709704, + "loss": 2.4994, "step": 451 }, { - "epoch": 0.14475580464371496, - "grad_norm": 0.55078125, - "learning_rate": 4.969851822392805e-05, - "loss": 2.0273, + "epoch": 0.17561496806779825, + "grad_norm": 0.9375, + "learning_rate": 0.00019655442803451301, + "loss": 2.4471, "step": 452 }, { - "epoch": 0.14507606084867894, - "grad_norm": 0.470703125, - "learning_rate": 4.9694173650485595e-05, - "loss": 2.286, + "epoch": 0.17600349675821375, + "grad_norm": 0.95703125, + "learning_rate": 0.00019651902318244582, + "loss": 2.5041, "step": 453 }, { - "epoch": 0.14539631705364292, - "grad_norm": 0.640625, - "learning_rate": 4.968979818897694e-05, - "loss": 2.6001, + "epoch": 0.17639202544862922, + "grad_norm": 0.8359375, + "learning_rate": 0.0001964834405799707, + "loss": 2.4156, "step": 454 }, { - "epoch": 0.14571657325860687, - "grad_norm": 0.486328125, - "learning_rate": 4.968539184487502e-05, - "loss": 2.4443, + "epoch": 0.1767805541390447, + "grad_norm": 0.84375, + "learning_rate": 0.00019644768029261698, + "loss": 2.5044, "step": 455 }, { - "epoch": 0.14603682946357086, - "grad_norm": 0.62890625, - "learning_rate": 4.9680954623691374e-05, - "loss": 2.2168, + "epoch": 0.1771690828294602, + "grad_norm": 0.890625, + "learning_rate": 0.00019641174238624115, + "loss": 2.482, "step": 456 }, { - "epoch": 0.14635708566853484, - "grad_norm": 0.5, - "learning_rate": 4.9676486530976196e-05, - "loss": 2.4374, + "epoch": 0.17755761151987567, + "grad_norm": 0.8515625, + "learning_rate": 0.00019637562692702677, + "loss": 2.4772, "step": 457 }, { - "epoch": 0.1466773418734988, - "grad_norm": 0.46875, - "learning_rate": 4.967198757231829e-05, - "loss": 2.1691, + "epoch": 0.17794614021029115, + "grad_norm": 0.96484375, + "learning_rate": 0.00019633933398148452, + "loss": 2.3809, "step": 458 }, { - "epoch": 0.14699759807846277, - "grad_norm": 0.83203125, - "learning_rate": 4.966745775334505e-05, - "loss": 2.7222, + "epoch": 0.17833466890070665, + "grad_norm": 0.80078125, + "learning_rate": 0.0001963028636164518, + "loss": 2.4452, "step": 459 }, { - "epoch": 0.14731785428342675, - "grad_norm": 0.53125, - "learning_rate": 4.966289707972249e-05, - "loss": 1.9953, + "epoch": 0.17872319759112212, + "grad_norm": 1.015625, + "learning_rate": 0.0001962662158990928, + "loss": 2.4653, "step": 460 }, { - "epoch": 0.1476381104883907, - "grad_norm": 0.255859375, - "learning_rate": 4.965830555715522e-05, - "loss": 2.0458, + "epoch": 0.1791117262815376, + "grad_norm": 0.921875, + "learning_rate": 0.00019622939089689837, + "loss": 2.4565, "step": 461 }, { - "epoch": 0.1479583666933547, - "grad_norm": 0.56640625, - "learning_rate": 4.9653683191386405e-05, - "loss": 2.2034, + "epoch": 0.17950025497195307, + "grad_norm": 0.890625, + "learning_rate": 0.00019619238867768577, + "loss": 2.4595, "step": 462 }, { - "epoch": 0.14827862289831867, - "grad_norm": 0.470703125, - "learning_rate": 4.964902998819782e-05, - "loss": 2.0623, + "epoch": 0.17988878366236857, + "grad_norm": 0.9375, + "learning_rate": 0.00019615520930959867, + "loss": 2.4754, "step": 463 }, { - "epoch": 0.14859887910328262, - "grad_norm": 0.490234375, - "learning_rate": 4.964434595340982e-05, - "loss": 2.0543, + "epoch": 0.18027731235278405, + "grad_norm": 0.8359375, + "learning_rate": 0.00019611785286110695, + "loss": 2.4152, "step": 464 }, { - "epoch": 0.1489191353082466, - "grad_norm": 0.6171875, - "learning_rate": 4.963963109288129e-05, - "loss": 1.7578, + "epoch": 0.18066584104319952, + "grad_norm": 0.90234375, + "learning_rate": 0.00019608031940100665, + "loss": 2.4357, "step": 465 }, { - "epoch": 0.14923939151321056, - "grad_norm": 0.3515625, - "learning_rate": 4.963488541250969e-05, - "loss": 1.9211, + "epoch": 0.18105436973361502, + "grad_norm": 0.94921875, + "learning_rate": 0.00019604260899841979, + "loss": 2.5019, "step": 466 }, { - "epoch": 0.14955964771817454, - "grad_norm": 0.423828125, - "learning_rate": 4.963010891823103e-05, - "loss": 2.1433, + "epoch": 0.1814428984240305, + "grad_norm": 0.8984375, + "learning_rate": 0.0001960047217227942, + "loss": 2.5252, "step": 467 }, { - "epoch": 0.14987990392313852, - "grad_norm": 0.86328125, - "learning_rate": 4.962530161601988e-05, - "loss": 2.7066, + "epoch": 0.18183142711444597, + "grad_norm": 0.83203125, + "learning_rate": 0.00019596665764390342, + "loss": 2.3954, "step": 468 }, { - "epoch": 0.15020016012810247, - "grad_norm": 0.40234375, - "learning_rate": 4.9620463511889314e-05, - "loss": 1.7484, + "epoch": 0.18221995580486147, + "grad_norm": 1.03125, + "learning_rate": 0.0001959284168318467, + "loss": 2.4657, "step": 469 }, { - "epoch": 0.15052041633306645, - "grad_norm": 0.380859375, - "learning_rate": 4.961559461189096e-05, - "loss": 2.2189, + "epoch": 0.18260848449527695, + "grad_norm": 0.9140625, + "learning_rate": 0.00019588999935704876, + "loss": 2.5376, "step": 470 }, { - "epoch": 0.15084067253803043, - "grad_norm": 0.33203125, - "learning_rate": 4.961069492211494e-05, - "loss": 1.9058, + "epoch": 0.18299701318569242, + "grad_norm": 0.88671875, + "learning_rate": 0.0001958514052902595, + "loss": 2.3757, "step": 471 }, { - "epoch": 0.1511609287429944, - "grad_norm": 0.490234375, - "learning_rate": 4.960576444868992e-05, - "loss": 2.1691, + "epoch": 0.18338554187610792, + "grad_norm": 0.9921875, + "learning_rate": 0.00019581263470255417, + "loss": 2.4966, "step": 472 }, { - "epoch": 0.15148118494795837, - "grad_norm": 0.35546875, - "learning_rate": 4.960080319778305e-05, - "loss": 2.2587, + "epoch": 0.1837740705665234, + "grad_norm": 0.8515625, + "learning_rate": 0.0001957736876653331, + "loss": 2.4341, "step": 473 }, { - "epoch": 0.15180144115292235, - "grad_norm": 0.57421875, - "learning_rate": 4.959581117559998e-05, - "loss": 1.6996, + "epoch": 0.18416259925693887, + "grad_norm": 0.84375, + "learning_rate": 0.00019573456425032158, + "loss": 2.3674, "step": 474 }, { - "epoch": 0.1521216973578863, - "grad_norm": 0.419921875, - "learning_rate": 4.959078838838485e-05, - "loss": 2.3875, + "epoch": 0.18455112794735437, + "grad_norm": 1.0234375, + "learning_rate": 0.00019569526452956962, + "loss": 2.4213, "step": 475 }, { - "epoch": 0.15244195356285029, - "grad_norm": 0.515625, - "learning_rate": 4.958573484242027e-05, - "loss": 2.4589, + "epoch": 0.18493965663776984, + "grad_norm": 0.90625, + "learning_rate": 0.00019565578857545202, + "loss": 2.4737, "step": 476 }, { - "epoch": 0.15276220976781424, - "grad_norm": 2.578125, - "learning_rate": 4.9580650544027376e-05, - "loss": 2.1532, + "epoch": 0.18532818532818532, + "grad_norm": 0.8515625, + "learning_rate": 0.0001956161364606681, + "loss": 2.4017, "step": 477 }, { - "epoch": 0.15308246597277822, - "grad_norm": 0.478515625, - "learning_rate": 4.957553549956568e-05, - "loss": 2.2743, + "epoch": 0.18571671401860082, + "grad_norm": 0.90234375, + "learning_rate": 0.00019557630825824156, + "loss": 2.515, "step": 478 }, { - "epoch": 0.1534027221777422, - "grad_norm": 0.435546875, - "learning_rate": 4.957038971543325e-05, - "loss": 1.9105, + "epoch": 0.1861052427090163, + "grad_norm": 1.0703125, + "learning_rate": 0.00019553630404152043, + "loss": 2.4351, "step": 479 }, { - "epoch": 0.15372297838270615, - "grad_norm": 0.376953125, - "learning_rate": 4.956521319806653e-05, - "loss": 2.2667, + "epoch": 0.18649377139943177, + "grad_norm": 0.87890625, + "learning_rate": 0.00019549612388417692, + "loss": 2.3931, "step": 480 }, { - "epoch": 0.15404323458767014, - "grad_norm": 0.5625, - "learning_rate": 4.956000595394043e-05, - "loss": 1.9691, + "epoch": 0.18688230008984727, + "grad_norm": 0.94921875, + "learning_rate": 0.00019545576786020717, + "loss": 2.4537, "step": 481 }, { - "epoch": 0.15436349079263412, - "grad_norm": 1.03125, - "learning_rate": 4.955476798956831e-05, - "loss": 1.549, + "epoch": 0.18727082878026274, + "grad_norm": 0.87890625, + "learning_rate": 0.00019541523604393128, + "loss": 2.514, "step": 482 }, { - "epoch": 0.15468374699759807, - "grad_norm": 0.50390625, - "learning_rate": 4.954949931150193e-05, - "loss": 1.8682, + "epoch": 0.18765935747067822, + "grad_norm": 0.91796875, + "learning_rate": 0.00019537452850999302, + "loss": 2.4735, "step": 483 }, { - "epoch": 0.15500400320256205, - "grad_norm": 0.51171875, - "learning_rate": 4.9544199926331484e-05, - "loss": 1.846, + "epoch": 0.18804788616109372, + "grad_norm": 0.875, + "learning_rate": 0.0001953336453333598, + "loss": 2.4783, "step": 484 }, { - "epoch": 0.15532425940752603, - "grad_norm": 0.66796875, - "learning_rate": 4.953886984068558e-05, - "loss": 2.1578, + "epoch": 0.1884364148515092, + "grad_norm": 1.03125, + "learning_rate": 0.00019529258658932248, + "loss": 2.4402, "step": 485 }, { - "epoch": 0.15564451561249, - "grad_norm": 0.546875, - "learning_rate": 4.95335090612312e-05, - "loss": 2.2433, + "epoch": 0.18882494354192467, + "grad_norm": 0.88671875, + "learning_rate": 0.0001952513523534953, + "loss": 2.446, "step": 486 }, { - "epoch": 0.15596477181745397, - "grad_norm": 0.58984375, - "learning_rate": 4.952811759467374e-05, - "loss": 2.2284, + "epoch": 0.18921347223234017, + "grad_norm": 0.87890625, + "learning_rate": 0.0001952099427018156, + "loss": 2.4653, "step": 487 }, { - "epoch": 0.15628502802241792, - "grad_norm": 0.478515625, - "learning_rate": 4.9522695447757e-05, - "loss": 2.5033, + "epoch": 0.18960200092275564, + "grad_norm": 0.9296875, + "learning_rate": 0.00019516835771054382, + "loss": 2.439, "step": 488 }, { - "epoch": 0.1566052842273819, - "grad_norm": 0.39453125, - "learning_rate": 4.9517242627263096e-05, - "loss": 1.7995, + "epoch": 0.18999052961317112, + "grad_norm": 0.86328125, + "learning_rate": 0.00019512659745626334, + "loss": 2.5075, "step": 489 }, { - "epoch": 0.15692554043234588, - "grad_norm": 0.4375, - "learning_rate": 4.951175914001256e-05, - "loss": 2.5606, + "epoch": 0.19037905830358662, + "grad_norm": 0.83203125, + "learning_rate": 0.0001950846620158802, + "loss": 2.4529, "step": 490 }, { - "epoch": 0.15724579663730984, - "grad_norm": 1.7265625, - "learning_rate": 4.950624499286428e-05, - "loss": 2.3475, + "epoch": 0.1907675869940021, + "grad_norm": 0.83203125, + "learning_rate": 0.00019504255146662321, + "loss": 2.3809, "step": 491 }, { - "epoch": 0.15756605284227382, - "grad_norm": 0.474609375, - "learning_rate": 4.950070019271547e-05, - "loss": 2.3772, + "epoch": 0.19115611568441757, + "grad_norm": 0.859375, + "learning_rate": 0.0001950002658860435, + "loss": 2.4723, "step": 492 }, { - "epoch": 0.1578863090472378, - "grad_norm": 0.40234375, - "learning_rate": 4.949512474650171e-05, - "loss": 1.9745, + "epoch": 0.19154464437483307, + "grad_norm": 0.890625, + "learning_rate": 0.0001949578053520147, + "loss": 2.4074, "step": 493 }, { - "epoch": 0.15820656525220175, - "grad_norm": 0.58984375, - "learning_rate": 4.94895186611969e-05, - "loss": 2.5581, + "epoch": 0.19193317306524854, + "grad_norm": 0.88671875, + "learning_rate": 0.0001949151699427325, + "loss": 2.4241, "step": 494 }, { - "epoch": 0.15852682145716573, - "grad_norm": 0.82421875, - "learning_rate": 4.948388194381326e-05, - "loss": 2.3336, + "epoch": 0.19232170175566402, + "grad_norm": 0.859375, + "learning_rate": 0.00019487235973671471, + "loss": 2.4672, "step": 495 }, { - "epoch": 0.15884707766212972, - "grad_norm": 0.443359375, - "learning_rate": 4.947821460140134e-05, - "loss": 2.1051, + "epoch": 0.1927102304460795, + "grad_norm": 0.87109375, + "learning_rate": 0.0001948293748128011, + "loss": 2.4237, "step": 496 }, { - "epoch": 0.15916733386709367, - "grad_norm": 0.55078125, - "learning_rate": 4.947251664104998e-05, - "loss": 2.4866, + "epoch": 0.193098759136495, + "grad_norm": 0.87890625, + "learning_rate": 0.00019478621525015303, + "loss": 2.442, "step": 497 }, { - "epoch": 0.15948759007205765, - "grad_norm": 0.419921875, - "learning_rate": 4.946678806988633e-05, - "loss": 2.2679, + "epoch": 0.19348728782691046, + "grad_norm": 0.94921875, + "learning_rate": 0.00019474288112825367, + "loss": 2.4969, "step": 498 }, { - "epoch": 0.1598078462770216, - "grad_norm": 0.419921875, - "learning_rate": 4.9461028895075825e-05, - "loss": 2.4515, + "epoch": 0.19387581651732594, + "grad_norm": 0.86328125, + "learning_rate": 0.00019469937252690754, + "loss": 2.4462, "step": 499 }, { - "epoch": 0.16012810248198558, - "grad_norm": 0.466796875, - "learning_rate": 4.9455239123822184e-05, - "loss": 1.6854, + "epoch": 0.19426434520774144, + "grad_norm": 0.94140625, + "learning_rate": 0.00019465568952624053, + "loss": 2.3824, "step": 500 }, { - "epoch": 0.16044835868694957, - "grad_norm": 0.97265625, - "learning_rate": 4.944941876336738e-05, - "loss": 2.0233, + "epoch": 0.19465287389815691, + "grad_norm": 0.984375, + "learning_rate": 0.00019461183220669974, + "loss": 2.5187, "step": 501 }, { - "epoch": 0.16076861489191352, - "grad_norm": 0.5390625, - "learning_rate": 4.944356782099167e-05, - "loss": 2.467, + "epoch": 0.1950414025885724, + "grad_norm": 0.921875, + "learning_rate": 0.00019456780064905317, + "loss": 2.4442, "step": 502 }, { - "epoch": 0.1610888710968775, - "grad_norm": 0.4921875, - "learning_rate": 4.943768630401355e-05, - "loss": 2.4588, + "epoch": 0.1954299312789879, + "grad_norm": 1.0234375, + "learning_rate": 0.00019452359493438983, + "loss": 2.4223, "step": 503 }, { - "epoch": 0.16140912730184148, - "grad_norm": 0.5625, - "learning_rate": 4.943177421978976e-05, - "loss": 2.4571, + "epoch": 0.19581845996940336, + "grad_norm": 0.83203125, + "learning_rate": 0.0001944792151441194, + "loss": 2.544, "step": 504 }, { - "epoch": 0.16172938350680544, - "grad_norm": 0.408203125, - "learning_rate": 4.9425831575715286e-05, - "loss": 2.6707, + "epoch": 0.19620698865981884, + "grad_norm": 0.98046875, + "learning_rate": 0.00019443466135997216, + "loss": 2.4107, "step": 505 }, { - "epoch": 0.16204963971176942, - "grad_norm": 0.87890625, - "learning_rate": 4.941985837922332e-05, - "loss": 2.1903, + "epoch": 0.19659551735023434, + "grad_norm": 0.9921875, + "learning_rate": 0.0001943899336639988, + "loss": 2.4845, "step": 506 }, { - "epoch": 0.1623698959167334, - "grad_norm": 0.302734375, - "learning_rate": 4.941385463778528e-05, - "loss": 1.7641, + "epoch": 0.1969840460406498, + "grad_norm": 0.80078125, + "learning_rate": 0.00019434503213857027, + "loss": 2.4167, "step": 507 }, { - "epoch": 0.16269015212169735, - "grad_norm": 0.89453125, - "learning_rate": 4.9407820358910804e-05, - "loss": 2.4916, + "epoch": 0.1973725747310653, + "grad_norm": 0.98828125, + "learning_rate": 0.0001942999568663777, + "loss": 2.4629, "step": 508 }, { - "epoch": 0.16301040832666133, - "grad_norm": 1.59375, - "learning_rate": 4.94017555501477e-05, - "loss": 2.0425, + "epoch": 0.1977611034214808, + "grad_norm": 0.94140625, + "learning_rate": 0.0001942547079304322, + "loss": 2.4738, "step": 509 }, { - "epoch": 0.1633306645316253, - "grad_norm": 0.48046875, - "learning_rate": 4.939566021908197e-05, - "loss": 2.0939, + "epoch": 0.19814963211189626, + "grad_norm": 0.94921875, + "learning_rate": 0.00019420928541406463, + "loss": 2.4505, "step": 510 }, { - "epoch": 0.16365092073658927, - "grad_norm": 0.609375, - "learning_rate": 4.938953437333783e-05, - "loss": 2.3065, + "epoch": 0.19853816080231174, + "grad_norm": 1.5390625, + "learning_rate": 0.00019416368940092555, + "loss": 2.4139, "step": 511 }, { - "epoch": 0.16397117694155325, - "grad_norm": 0.443359375, - "learning_rate": 4.93833780205776e-05, - "loss": 2.7856, + "epoch": 0.19892668949272724, + "grad_norm": 0.97265625, + "learning_rate": 0.00019411791997498502, + "loss": 2.4519, "step": 512 }, { - "epoch": 0.1642914331465172, - "grad_norm": 0.51953125, - "learning_rate": 4.937719116850181e-05, - "loss": 2.3894, + "epoch": 0.1993152181831427, + "grad_norm": 1.0546875, + "learning_rate": 0.00019407197722053255, + "loss": 2.4682, "step": 513 }, { - "epoch": 0.16461168935148118, - "grad_norm": 0.419921875, - "learning_rate": 4.937097382484913e-05, - "loss": 2.5719, + "epoch": 0.19970374687355819, + "grad_norm": 1.0546875, + "learning_rate": 0.0001940258612221767, + "loss": 2.4073, "step": 514 }, { - "epoch": 0.16493194555644516, - "grad_norm": 0.392578125, - "learning_rate": 4.936472599739635e-05, - "loss": 1.8058, + "epoch": 0.2000922755639737, + "grad_norm": 0.9609375, + "learning_rate": 0.0001939795720648452, + "loss": 2.4505, "step": 515 }, { - "epoch": 0.16525220176140912, - "grad_norm": 0.43359375, - "learning_rate": 4.935844769395842e-05, - "loss": 2.2764, + "epoch": 0.20048080425438916, + "grad_norm": 0.8984375, + "learning_rate": 0.00019393310983378462, + "loss": 2.4493, "step": 516 }, { - "epoch": 0.1655724579663731, - "grad_norm": 0.46875, - "learning_rate": 4.9352138922388375e-05, - "loss": 1.6205, + "epoch": 0.20086933294480463, + "grad_norm": 0.9921875, + "learning_rate": 0.0001938864746145603, + "loss": 2.4376, "step": 517 }, { - "epoch": 0.16589271417133708, - "grad_norm": 0.59765625, - "learning_rate": 4.9345799690577385e-05, - "loss": 1.8822, + "epoch": 0.20125786163522014, + "grad_norm": 1.046875, + "learning_rate": 0.00019383966649305607, + "loss": 2.3207, "step": 518 }, { - "epoch": 0.16621297037630103, - "grad_norm": 0.6953125, - "learning_rate": 4.933943000645471e-05, - "loss": 2.3701, + "epoch": 0.2016463903256356, + "grad_norm": 0.84765625, + "learning_rate": 0.00019379268555547432, + "loss": 2.522, "step": 519 }, { - "epoch": 0.16653322658126501, - "grad_norm": 1.2265625, - "learning_rate": 4.9333029877987715e-05, - "loss": 2.474, + "epoch": 0.20203491901605108, + "grad_norm": 0.921875, + "learning_rate": 0.00019374553188833554, + "loss": 2.491, "step": 520 }, { - "epoch": 0.166853482786229, - "grad_norm": 0.6015625, - "learning_rate": 4.932659931318182e-05, - "loss": 2.0294, + "epoch": 0.20242344770646659, + "grad_norm": 0.921875, + "learning_rate": 0.00019369820557847844, + "loss": 2.4827, "step": 521 }, { - "epoch": 0.16717373899119295, - "grad_norm": 0.46484375, - "learning_rate": 4.932013832008054e-05, - "loss": 2.218, + "epoch": 0.20281197639688206, + "grad_norm": 0.796875, + "learning_rate": 0.00019365070671305962, + "loss": 2.4586, "step": 522 }, { - "epoch": 0.16749399519615693, - "grad_norm": 0.828125, - "learning_rate": 4.931364690676544e-05, - "loss": 2.5899, + "epoch": 0.20320050508729753, + "grad_norm": 0.8515625, + "learning_rate": 0.0001936030353795535, + "loss": 2.4202, "step": 523 }, { - "epoch": 0.16781425140112088, - "grad_norm": 0.55859375, - "learning_rate": 4.930712508135613e-05, - "loss": 1.828, + "epoch": 0.20358903377771304, + "grad_norm": 0.90625, + "learning_rate": 0.00019355519166575205, + "loss": 2.4167, "step": 524 }, { - "epoch": 0.16813450760608487, - "grad_norm": 5.3125, - "learning_rate": 4.930057285201027e-05, - "loss": 2.627, + "epoch": 0.2039775624681285, + "grad_norm": 0.875, + "learning_rate": 0.00019350717565976474, + "loss": 2.4965, "step": 525 }, { - "epoch": 0.16845476381104885, - "grad_norm": 1.25, - "learning_rate": 4.929399022692355e-05, - "loss": 2.1664, + "epoch": 0.20436609115854398, + "grad_norm": 0.81640625, + "learning_rate": 0.00019345898745001838, + "loss": 2.4731, "step": 526 }, { - "epoch": 0.1687750200160128, - "grad_norm": 0.43359375, - "learning_rate": 4.928737721432967e-05, - "loss": 1.8216, + "epoch": 0.20475461984895948, + "grad_norm": 0.87890625, + "learning_rate": 0.0001934106271252568, + "loss": 2.4869, "step": 527 }, { - "epoch": 0.16909527622097678, - "grad_norm": 0.52734375, - "learning_rate": 4.9280733822500346e-05, - "loss": 1.9037, + "epoch": 0.20514314853937496, + "grad_norm": 0.83984375, + "learning_rate": 0.00019336209477454087, + "loss": 2.3862, "step": 528 }, { - "epoch": 0.16941553242594076, - "grad_norm": 0.48046875, - "learning_rate": 4.927406005974529e-05, - "loss": 2.0505, + "epoch": 0.20553167722979043, + "grad_norm": 0.8515625, + "learning_rate": 0.0001933133904872483, + "loss": 2.4865, "step": 529 }, { - "epoch": 0.16973578863090472, - "grad_norm": 0.64453125, - "learning_rate": 4.9267355934412214e-05, - "loss": 2.3799, + "epoch": 0.2059202059202059, + "grad_norm": 3.5, + "learning_rate": 0.0001932645143530734, + "loss": 2.4302, "step": 530 }, { - "epoch": 0.1700560448358687, - "grad_norm": 0.52734375, - "learning_rate": 4.926062145488679e-05, - "loss": 2.2806, + "epoch": 0.2063087346106214, + "grad_norm": 0.98828125, + "learning_rate": 0.00019321546646202688, + "loss": 2.4552, "step": 531 }, { - "epoch": 0.17037630104083268, - "grad_norm": 0.314453125, - "learning_rate": 4.9253856629592663e-05, - "loss": 2.4759, + "epoch": 0.20669726330103688, + "grad_norm": 0.83984375, + "learning_rate": 0.0001931662469044359, + "loss": 2.4428, "step": 532 }, { - "epoch": 0.17069655724579663, - "grad_norm": 0.486328125, - "learning_rate": 4.924706146699146e-05, - "loss": 2.6853, + "epoch": 0.20708579199145236, + "grad_norm": 0.80078125, + "learning_rate": 0.00019311685577094368, + "loss": 2.5089, "step": 533 }, { - "epoch": 0.1710168134507606, - "grad_norm": 0.5546875, - "learning_rate": 4.924023597558271e-05, - "loss": 1.9226, + "epoch": 0.20747432068186786, + "grad_norm": 0.80859375, + "learning_rate": 0.0001930672931525094, + "loss": 2.4275, "step": 534 }, { - "epoch": 0.17133706965572457, - "grad_norm": 0.8515625, - "learning_rate": 4.923338016390392e-05, - "loss": 1.8459, + "epoch": 0.20786284937228333, + "grad_norm": 0.7734375, + "learning_rate": 0.0001930175591404081, + "loss": 2.3744, "step": 535 }, { - "epoch": 0.17165732586068855, - "grad_norm": 0.408203125, - "learning_rate": 4.922649404053048e-05, - "loss": 2.0748, + "epoch": 0.2082513780626988, + "grad_norm": 0.875, + "learning_rate": 0.00019296765382623049, + "loss": 2.4922, "step": 536 }, { - "epoch": 0.17197758206565253, - "grad_norm": 0.72265625, - "learning_rate": 4.921957761407575e-05, - "loss": 2.0169, + "epoch": 0.2086399067531143, + "grad_norm": 0.8359375, + "learning_rate": 0.0001929175773018826, + "loss": 2.5342, "step": 537 }, { - "epoch": 0.17229783827061648, - "grad_norm": 0.75390625, - "learning_rate": 4.921263089319095e-05, - "loss": 1.7651, + "epoch": 0.20902843544352978, + "grad_norm": 1.0234375, + "learning_rate": 0.00019286732965958592, + "loss": 2.4245, "step": 538 }, { - "epoch": 0.17261809447558046, - "grad_norm": 0.53125, - "learning_rate": 4.920565388656519e-05, - "loss": 2.498, + "epoch": 0.20941696413394525, + "grad_norm": 0.828125, + "learning_rate": 0.00019281691099187702, + "loss": 2.4321, "step": 539 }, { - "epoch": 0.17293835068054444, - "grad_norm": 0.52734375, - "learning_rate": 4.919864660292549e-05, - "loss": 2.1956, + "epoch": 0.20980549282436076, + "grad_norm": 1.09375, + "learning_rate": 0.0001927663213916074, + "loss": 2.4336, "step": 540 }, { - "epoch": 0.1732586068855084, - "grad_norm": 0.671875, - "learning_rate": 4.919160905103673e-05, - "loss": 2.4898, + "epoch": 0.21019402151477623, + "grad_norm": 1.0546875, + "learning_rate": 0.0001927155609519434, + "loss": 2.4085, "step": 541 }, { - "epoch": 0.17357886309047238, - "grad_norm": 0.494140625, - "learning_rate": 4.9184541239701666e-05, - "loss": 2.1794, + "epoch": 0.2105825502051917, + "grad_norm": 1.109375, + "learning_rate": 0.00019266462976636598, + "loss": 2.4764, "step": 542 }, { - "epoch": 0.17389911929543636, - "grad_norm": 0.5, - "learning_rate": 4.9177443177760854e-05, - "loss": 2.1124, + "epoch": 0.2109710788956072, + "grad_norm": 0.9296875, + "learning_rate": 0.00019261352792867046, + "loss": 2.4533, "step": 543 }, { - "epoch": 0.17421937550040031, - "grad_norm": 1.0078125, - "learning_rate": 4.9170314874092734e-05, - "loss": 2.161, + "epoch": 0.21135960758602268, + "grad_norm": 0.82421875, + "learning_rate": 0.00019256225553296655, + "loss": 2.4508, "step": 544 }, { - "epoch": 0.1745396317053643, - "grad_norm": 0.66796875, - "learning_rate": 4.916315633761356e-05, - "loss": 2.2314, + "epoch": 0.21174813627643815, + "grad_norm": 0.8671875, + "learning_rate": 0.00019251081267367807, + "loss": 2.3913, "step": 545 }, { - "epoch": 0.17485988791032825, - "grad_norm": 0.400390625, - "learning_rate": 4.915596757727741e-05, - "loss": 1.9608, + "epoch": 0.21213666496685366, + "grad_norm": 0.87890625, + "learning_rate": 0.00019245919944554257, + "loss": 2.3765, "step": 546 }, { - "epoch": 0.17518014411529223, - "grad_norm": 0.46484375, - "learning_rate": 4.9148748602076146e-05, - "loss": 1.8623, + "epoch": 0.21252519365726913, + "grad_norm": 0.953125, + "learning_rate": 0.00019240741594361155, + "loss": 2.3952, "step": 547 }, { - "epoch": 0.1755004003202562, - "grad_norm": 0.5078125, - "learning_rate": 4.914149942103944e-05, - "loss": 2.0954, + "epoch": 0.2129137223476846, + "grad_norm": 0.8984375, + "learning_rate": 0.0001923554622632501, + "loss": 2.4084, "step": 548 }, { - "epoch": 0.17582065652522016, - "grad_norm": 0.36328125, - "learning_rate": 4.913422004323474e-05, - "loss": 2.0595, + "epoch": 0.2133022510381001, + "grad_norm": 0.953125, + "learning_rate": 0.00019230333850013654, + "loss": 2.3587, "step": 549 }, { - "epoch": 0.17614091273018415, - "grad_norm": 0.369140625, - "learning_rate": 4.912691047776727e-05, - "loss": 2.1955, + "epoch": 0.21369077972851558, + "grad_norm": 1.0625, + "learning_rate": 0.00019225104475026258, + "loss": 2.3831, "step": 550 }, { - "epoch": 0.17646116893514813, - "grad_norm": 0.470703125, - "learning_rate": 4.9119570733780015e-05, - "loss": 2.0215, + "epoch": 0.21407930841893105, + "grad_norm": 0.85546875, + "learning_rate": 0.00019219858110993288, + "loss": 2.3703, "step": 551 }, { - "epoch": 0.17678142514011208, - "grad_norm": 0.498046875, - "learning_rate": 4.91122008204537e-05, - "loss": 2.29, + "epoch": 0.21446783710934655, + "grad_norm": 1.0078125, + "learning_rate": 0.00019214594767576508, + "loss": 2.4301, "step": 552 }, { - "epoch": 0.17710168134507606, - "grad_norm": 0.494140625, - "learning_rate": 4.910480074700678e-05, - "loss": 2.1564, + "epoch": 0.21485636579976203, + "grad_norm": 1.15625, + "learning_rate": 0.00019209314454468937, + "loss": 2.4783, "step": 553 }, { - "epoch": 0.17742193755004004, - "grad_norm": 0.47265625, - "learning_rate": 4.909737052269545e-05, - "loss": 2.0404, + "epoch": 0.2152448944901775, + "grad_norm": 0.8515625, + "learning_rate": 0.00019204017181394855, + "loss": 2.4635, "step": 554 }, { - "epoch": 0.177742193755004, - "grad_norm": 0.4296875, - "learning_rate": 4.908991015681362e-05, - "loss": 2.3606, + "epoch": 0.215633423180593, + "grad_norm": 1.0625, + "learning_rate": 0.00019198702958109776, + "loss": 2.3172, "step": 555 }, { - "epoch": 0.17806244995996798, - "grad_norm": 0.6328125, - "learning_rate": 4.908241965869289e-05, - "loss": 2.4787, + "epoch": 0.21602195187100848, + "grad_norm": 0.9609375, + "learning_rate": 0.00019193371794400425, + "loss": 2.4102, "step": 556 }, { - "epoch": 0.17838270616493196, - "grad_norm": 0.48828125, - "learning_rate": 4.907489903770256e-05, - "loss": 2.3174, + "epoch": 0.21641048056142395, + "grad_norm": 0.8828125, + "learning_rate": 0.00019188023700084726, + "loss": 2.4023, "step": 557 }, { - "epoch": 0.1787029623698959, - "grad_norm": 0.484375, - "learning_rate": 4.9067348303249594e-05, - "loss": 2.5595, + "epoch": 0.21679900925183945, + "grad_norm": 0.890625, + "learning_rate": 0.00019182658685011785, + "loss": 2.5074, "step": 558 }, { - "epoch": 0.1790232185748599, - "grad_norm": 0.5625, - "learning_rate": 4.905976746477864e-05, - "loss": 2.6765, + "epoch": 0.21718753794225493, + "grad_norm": 0.8671875, + "learning_rate": 0.0001917727675906187, + "loss": 2.4028, "step": 559 }, { - "epoch": 0.17934347477982385, - "grad_norm": 0.490234375, - "learning_rate": 4.905215653177199e-05, - "loss": 2.6524, + "epoch": 0.2175760666326704, + "grad_norm": 0.82421875, + "learning_rate": 0.00019171877932146387, + "loss": 2.4564, "step": 560 }, { - "epoch": 0.17966373098478783, - "grad_norm": 0.5, - "learning_rate": 4.904451551374959e-05, - "loss": 2.497, + "epoch": 0.2179645953230859, + "grad_norm": 0.81640625, + "learning_rate": 0.00019166462214207869, + "loss": 2.4189, "step": 561 }, { - "epoch": 0.1799839871897518, - "grad_norm": 0.52734375, - "learning_rate": 4.903684442026899e-05, - "loss": 2.0149, + "epoch": 0.21835312401350138, + "grad_norm": 0.875, + "learning_rate": 0.00019161029615219962, + "loss": 2.4836, "step": 562 }, { - "epoch": 0.18030424339471576, - "grad_norm": 0.486328125, - "learning_rate": 4.9029143260925395e-05, - "loss": 2.1776, + "epoch": 0.21874165270391685, + "grad_norm": 0.84765625, + "learning_rate": 0.0001915558014518739, + "loss": 2.4663, "step": 563 }, { - "epoch": 0.18062449959967974, - "grad_norm": 0.470703125, - "learning_rate": 4.902141204535159e-05, - "loss": 1.8415, + "epoch": 0.21913018139433232, + "grad_norm": 0.8515625, + "learning_rate": 0.00019150113814145957, + "loss": 2.3867, "step": 564 }, { - "epoch": 0.18094475580464373, - "grad_norm": 0.53515625, - "learning_rate": 4.901365078321798e-05, - "loss": 2.2762, + "epoch": 0.21951871008474783, + "grad_norm": 0.8671875, + "learning_rate": 0.00019144630632162508, + "loss": 2.4438, "step": 565 }, { - "epoch": 0.18126501200960768, - "grad_norm": 0.40234375, - "learning_rate": 4.9005859484232526e-05, - "loss": 2.6546, + "epoch": 0.2199072387751633, + "grad_norm": 0.875, + "learning_rate": 0.0001913913060933493, + "loss": 2.4158, "step": 566 }, { - "epoch": 0.18158526821457166, - "grad_norm": 0.45703125, - "learning_rate": 4.899803815814077e-05, - "loss": 2.1712, + "epoch": 0.22029576746557877, + "grad_norm": 0.90625, + "learning_rate": 0.0001913361375579212, + "loss": 2.4088, "step": 567 }, { - "epoch": 0.18190552441953564, - "grad_norm": 0.400390625, - "learning_rate": 4.899018681472582e-05, - "loss": 2.5383, + "epoch": 0.22068429615599428, + "grad_norm": 0.97265625, + "learning_rate": 0.0001912808008169397, + "loss": 2.4266, "step": 568 }, { - "epoch": 0.1822257806244996, - "grad_norm": 0.76953125, - "learning_rate": 4.8982305463808334e-05, - "loss": 2.3668, + "epoch": 0.22107282484640975, + "grad_norm": 0.82421875, + "learning_rate": 0.00019122529597231352, + "loss": 2.4021, "step": 569 }, { - "epoch": 0.18254603682946358, - "grad_norm": 0.431640625, - "learning_rate": 4.897439411524648e-05, - "loss": 2.4888, + "epoch": 0.22146135353682522, + "grad_norm": 1.015625, + "learning_rate": 0.00019116962312626092, + "loss": 2.4939, "step": 570 }, { - "epoch": 0.18286629303442753, - "grad_norm": 0.376953125, - "learning_rate": 4.896645277893597e-05, - "loss": 1.5993, + "epoch": 0.22184988222724072, + "grad_norm": 1.0234375, + "learning_rate": 0.00019111378238130956, + "loss": 2.4856, "step": 571 }, { - "epoch": 0.1831865492393915, - "grad_norm": 0.50390625, - "learning_rate": 4.895848146481002e-05, - "loss": 2.1428, + "epoch": 0.2222384109176562, + "grad_norm": 5.90625, + "learning_rate": 0.0001910577738402964, + "loss": 2.5271, "step": 572 }, { - "epoch": 0.1835068054443555, - "grad_norm": 0.921875, - "learning_rate": 4.895048018283933e-05, - "loss": 2.6575, + "epoch": 0.22262693960807167, + "grad_norm": 1.0, + "learning_rate": 0.00019100159760636727, + "loss": 2.4331, "step": 573 }, { - "epoch": 0.18382706164931945, - "grad_norm": 1.9296875, - "learning_rate": 4.8942448943032116e-05, - "loss": 2.0452, + "epoch": 0.22301546829848717, + "grad_norm": 0.953125, + "learning_rate": 0.00019094525378297687, + "loss": 2.4777, "step": 574 }, { - "epoch": 0.18414731785428343, - "grad_norm": 0.57421875, - "learning_rate": 4.893438775543403e-05, - "loss": 1.9505, + "epoch": 0.22340399698890265, + "grad_norm": 0.828125, + "learning_rate": 0.00019088874247388864, + "loss": 2.5293, "step": 575 }, { - "epoch": 0.1844675740592474, - "grad_norm": 0.439453125, - "learning_rate": 4.89262966301282e-05, - "loss": 1.9657, + "epoch": 0.22379252567931812, + "grad_norm": 0.8515625, + "learning_rate": 0.0001908320637831743, + "loss": 2.4537, "step": 576 }, { - "epoch": 0.18478783026421136, - "grad_norm": 0.46875, - "learning_rate": 4.8918175577235214e-05, - "loss": 2.5377, + "epoch": 0.22418105436973362, + "grad_norm": 1.0390625, + "learning_rate": 0.00019077521781521388, + "loss": 2.4579, "step": 577 }, { - "epoch": 0.18510808646917534, - "grad_norm": 0.40625, - "learning_rate": 4.891002460691306e-05, - "loss": 2.1608, + "epoch": 0.2245695830601491, + "grad_norm": 0.9140625, + "learning_rate": 0.00019071820467469558, + "loss": 2.4129, "step": 578 }, { - "epoch": 0.18542834267413932, - "grad_norm": 0.421875, - "learning_rate": 4.890184372935719e-05, - "loss": 2.0481, + "epoch": 0.22495811175056457, + "grad_norm": 0.9140625, + "learning_rate": 0.0001906610244666153, + "loss": 2.3975, "step": 579 }, { - "epoch": 0.18574859887910328, - "grad_norm": 0.404296875, - "learning_rate": 4.889363295480042e-05, - "loss": 1.9898, + "epoch": 0.22534664044098007, + "grad_norm": 0.8671875, + "learning_rate": 0.00019060367729627673, + "loss": 2.4488, "step": 580 }, { - "epoch": 0.18606885508406726, - "grad_norm": 0.5859375, - "learning_rate": 4.8885392293513e-05, - "loss": 2.3255, + "epoch": 0.22573516913139555, + "grad_norm": 0.89453125, + "learning_rate": 0.000190546163269291, + "loss": 2.5263, "step": 581 }, { - "epoch": 0.1863891112890312, - "grad_norm": 0.6171875, - "learning_rate": 4.8877121755802536e-05, - "loss": 2.2675, + "epoch": 0.22612369782181102, + "grad_norm": 0.8359375, + "learning_rate": 0.00019048848249157647, + "loss": 2.4725, "step": 582 }, { - "epoch": 0.1867093674939952, - "grad_norm": 0.60546875, - "learning_rate": 4.8868821352014024e-05, - "loss": 2.2714, + "epoch": 0.22651222651222652, + "grad_norm": 0.8359375, + "learning_rate": 0.0001904306350693587, + "loss": 2.4523, "step": 583 }, { - "epoch": 0.18702962369895917, - "grad_norm": 0.5703125, - "learning_rate": 4.886049109252981e-05, - "loss": 2.0052, + "epoch": 0.226900755202642, + "grad_norm": 0.79296875, + "learning_rate": 0.00019037262110917008, + "loss": 2.5423, "step": 584 }, { - "epoch": 0.18734987990392313, - "grad_norm": 0.8125, - "learning_rate": 4.8852130987769574e-05, - "loss": 2.4493, + "epoch": 0.22728928389305747, + "grad_norm": 0.8203125, + "learning_rate": 0.00019031444071784966, + "loss": 2.4369, "step": 585 }, { - "epoch": 0.1876701361088871, - "grad_norm": 0.6953125, - "learning_rate": 4.884374104819035e-05, - "loss": 2.2796, + "epoch": 0.22767781258347297, + "grad_norm": 0.875, + "learning_rate": 0.00019025609400254308, + "loss": 2.4778, "step": 586 }, { - "epoch": 0.1879903923138511, - "grad_norm": 0.265625, - "learning_rate": 4.883532128428645e-05, - "loss": 1.4809, + "epoch": 0.22806634127388845, + "grad_norm": 0.8203125, + "learning_rate": 0.00019019758107070224, + "loss": 2.3663, "step": 587 }, { - "epoch": 0.18831064851881504, - "grad_norm": 0.359375, - "learning_rate": 4.882687170658955e-05, - "loss": 1.5241, + "epoch": 0.22845486996430392, + "grad_norm": 0.83203125, + "learning_rate": 0.00019013890203008514, + "loss": 2.3942, "step": 588 }, { - "epoch": 0.18863090472377902, - "grad_norm": 0.337890625, - "learning_rate": 4.881839232566856e-05, - "loss": 2.1891, + "epoch": 0.22884339865471942, + "grad_norm": 1.0078125, + "learning_rate": 0.00019008005698875567, + "loss": 2.4601, "step": 589 }, { - "epoch": 0.188951160928743, - "grad_norm": 0.640625, - "learning_rate": 4.880988315212971e-05, - "loss": 2.2043, + "epoch": 0.2292319273451349, + "grad_norm": 0.85546875, + "learning_rate": 0.00019002104605508343, + "loss": 2.4136, "step": 590 }, { - "epoch": 0.18927141713370696, - "grad_norm": 0.7265625, - "learning_rate": 4.880134419661647e-05, - "loss": 2.3487, + "epoch": 0.22962045603555037, + "grad_norm": 0.8671875, + "learning_rate": 0.00018996186933774366, + "loss": 2.4091, "step": 591 }, { - "epoch": 0.18959167333867094, - "grad_norm": 0.498046875, - "learning_rate": 4.879277546980957e-05, - "loss": 2.6678, + "epoch": 0.23000898472596587, + "grad_norm": 0.8671875, + "learning_rate": 0.00018990252694571667, + "loss": 2.3836, "step": 592 }, { - "epoch": 0.1899119295436349, - "grad_norm": 0.4375, - "learning_rate": 4.8784176982426984e-05, - "loss": 2.1424, + "epoch": 0.23039751341638134, + "grad_norm": 0.828125, + "learning_rate": 0.00018984301898828813, + "loss": 2.4327, "step": 593 }, { - "epoch": 0.19023218574859888, - "grad_norm": 0.455078125, - "learning_rate": 4.87755487452239e-05, - "loss": 2.3376, + "epoch": 0.23078604210679682, + "grad_norm": 0.87109375, + "learning_rate": 0.00018978334557504843, + "loss": 2.4264, "step": 594 }, { - "epoch": 0.19055244195356286, - "grad_norm": 0.46875, - "learning_rate": 4.8766890768992734e-05, - "loss": 2.4265, + "epoch": 0.23117457079721232, + "grad_norm": 0.8515625, + "learning_rate": 0.0001897235068158927, + "loss": 2.4683, "step": 595 }, { - "epoch": 0.1908726981585268, - "grad_norm": 0.65234375, - "learning_rate": 4.875820306456309e-05, - "loss": 2.366, + "epoch": 0.2315630994876278, + "grad_norm": 0.8359375, + "learning_rate": 0.00018966350282102065, + "loss": 2.4395, "step": 596 }, { - "epoch": 0.1911929543634908, - "grad_norm": 0.4765625, - "learning_rate": 4.8749485642801754e-05, - "loss": 2.4942, + "epoch": 0.23195162817804327, + "grad_norm": 0.91796875, + "learning_rate": 0.00018960333370093626, + "loss": 2.4799, "step": 597 }, { - "epoch": 0.19151321056845477, - "grad_norm": 0.431640625, - "learning_rate": 4.874073851461268e-05, - "loss": 1.9429, + "epoch": 0.23234015686845874, + "grad_norm": 1.125, + "learning_rate": 0.00018954299956644755, + "loss": 2.4338, "step": 598 }, { - "epoch": 0.19183346677341873, - "grad_norm": 0.5625, - "learning_rate": 4.873196169093699e-05, - "loss": 2.1505, + "epoch": 0.23272868555887424, + "grad_norm": 0.90234375, + "learning_rate": 0.00018948250052866646, + "loss": 2.4708, "step": 599 }, { - "epoch": 0.1921537229783827, - "grad_norm": 0.8046875, - "learning_rate": 4.872315518275296e-05, - "loss": 2.373, + "epoch": 0.23311721424928972, + "grad_norm": 0.80078125, + "learning_rate": 0.0001894218366990087, + "loss": 2.4036, "step": 600 }, { - "epoch": 0.1924739791833467, - "grad_norm": 0.5703125, - "learning_rate": 4.871431900107597e-05, - "loss": 2.2588, + "epoch": 0.2335057429397052, + "grad_norm": 0.94140625, + "learning_rate": 0.00018936100818919328, + "loss": 2.4747, "step": 601 }, { - "epoch": 0.19279423538831064, - "grad_norm": 0.71484375, - "learning_rate": 4.870545315695853e-05, - "loss": 2.6125, + "epoch": 0.2338942716301207, + "grad_norm": 1.078125, + "learning_rate": 0.0001893000151112427, + "loss": 2.5058, "step": 602 }, { - "epoch": 0.19311449159327462, - "grad_norm": 0.515625, - "learning_rate": 4.869655766149024e-05, - "loss": 2.489, + "epoch": 0.23428280032053617, + "grad_norm": 1.2265625, + "learning_rate": 0.00018923885757748238, + "loss": 2.4404, "step": 603 }, { - "epoch": 0.1934347477982386, - "grad_norm": 0.39453125, - "learning_rate": 4.868763252579782e-05, - "loss": 2.29, + "epoch": 0.23467132901095164, + "grad_norm": 0.94140625, + "learning_rate": 0.00018917753570054066, + "loss": 2.4196, "step": 604 }, { - "epoch": 0.19375500400320256, - "grad_norm": 0.51953125, - "learning_rate": 4.867867776104503e-05, - "loss": 2.3659, + "epoch": 0.23505985770136714, + "grad_norm": 0.8359375, + "learning_rate": 0.00018911604959334858, + "loss": 2.4353, "step": 605 }, { - "epoch": 0.19407526020816654, - "grad_norm": 0.427734375, - "learning_rate": 4.866969337843271e-05, - "loss": 1.8565, + "epoch": 0.23544838639178262, + "grad_norm": 0.8203125, + "learning_rate": 0.00018905439936913948, + "loss": 2.476, "step": 606 }, { - "epoch": 0.1943955164131305, - "grad_norm": 0.439453125, - "learning_rate": 4.8660679389198746e-05, - "loss": 2.5902, + "epoch": 0.2358369150821981, + "grad_norm": 1.0078125, + "learning_rate": 0.00018899258514144907, + "loss": 2.4251, "step": 607 }, { - "epoch": 0.19471577261809447, - "grad_norm": 0.365234375, - "learning_rate": 4.8651635804618046e-05, - "loss": 2.3034, + "epoch": 0.2362254437726136, + "grad_norm": 0.8359375, + "learning_rate": 0.00018893060702411508, + "loss": 2.4725, "step": 608 }, { - "epoch": 0.19503602882305845, - "grad_norm": 0.35546875, - "learning_rate": 4.8642562636002546e-05, - "loss": 2.4728, + "epoch": 0.23661397246302907, + "grad_norm": 0.86328125, + "learning_rate": 0.000188868465131277, + "loss": 2.4095, "step": 609 }, { - "epoch": 0.1953562850280224, - "grad_norm": 0.455078125, - "learning_rate": 4.8633459894701186e-05, - "loss": 2.1875, + "epoch": 0.23700250115344454, + "grad_norm": 0.80859375, + "learning_rate": 0.00018880615957737602, + "loss": 2.4845, "step": 610 }, { - "epoch": 0.1956765412329864, - "grad_norm": 0.466796875, - "learning_rate": 4.8624327592099884e-05, - "loss": 2.2461, + "epoch": 0.23739102984386004, + "grad_norm": 0.85546875, + "learning_rate": 0.00018874369047715465, + "loss": 2.3689, "step": 611 }, { - "epoch": 0.19599679743795037, - "grad_norm": 0.390625, - "learning_rate": 4.861516573962156e-05, - "loss": 1.8182, + "epoch": 0.23777955853427551, + "grad_norm": 0.77734375, + "learning_rate": 0.0001886810579456566, + "loss": 2.4247, "step": 612 }, { - "epoch": 0.19631705364291432, - "grad_norm": 0.99609375, - "learning_rate": 4.860597434872606e-05, - "loss": 1.7308, + "epoch": 0.238168087224691, + "grad_norm": 0.79296875, + "learning_rate": 0.00018861826209822663, + "loss": 2.3775, "step": 613 }, { - "epoch": 0.1966373098478783, - "grad_norm": 0.412109375, - "learning_rate": 4.859675343091021e-05, - "loss": 2.5587, + "epoch": 0.2385566159151065, + "grad_norm": 0.90234375, + "learning_rate": 0.00018855530305051011, + "loss": 2.4572, "step": 614 }, { - "epoch": 0.1969575660528423, - "grad_norm": 3.15625, - "learning_rate": 4.858750299770774e-05, - "loss": 2.2981, + "epoch": 0.23894514460552196, + "grad_norm": 0.85546875, + "learning_rate": 0.00018849218091845314, + "loss": 2.4004, "step": 615 }, { - "epoch": 0.19727782225780624, - "grad_norm": 0.984375, - "learning_rate": 4.8578223060689346e-05, - "loss": 2.356, + "epoch": 0.23933367329593744, + "grad_norm": 0.8125, + "learning_rate": 0.00018842889581830206, + "loss": 2.4783, "step": 616 }, { - "epoch": 0.19759807846277022, - "grad_norm": 0.47265625, - "learning_rate": 4.8568913631462565e-05, - "loss": 1.9542, + "epoch": 0.23972220198635294, + "grad_norm": 0.96875, + "learning_rate": 0.0001883654478666033, + "loss": 2.3635, "step": 617 }, { - "epoch": 0.19791833466773417, - "grad_norm": 0.369140625, - "learning_rate": 4.855957472167187e-05, - "loss": 2.3769, + "epoch": 0.2401107306767684, + "grad_norm": 0.91796875, + "learning_rate": 0.00018830183718020332, + "loss": 2.3983, "step": 618 }, { - "epoch": 0.19823859087269816, - "grad_norm": 0.56640625, - "learning_rate": 4.85502063429986e-05, - "loss": 2.4044, + "epoch": 0.2404992593671839, + "grad_norm": 0.8671875, + "learning_rate": 0.00018823806387624817, + "loss": 2.4074, "step": 619 }, { - "epoch": 0.19855884707766214, - "grad_norm": 0.53125, - "learning_rate": 4.8540808507160926e-05, - "loss": 2.4242, + "epoch": 0.2408877880575994, + "grad_norm": 1.1640625, + "learning_rate": 0.00018817412807218338, + "loss": 2.5352, "step": 620 }, { - "epoch": 0.1988791032826261, - "grad_norm": 0.65625, - "learning_rate": 4.85313812259139e-05, - "loss": 2.3783, + "epoch": 0.24127631674801486, + "grad_norm": 0.875, + "learning_rate": 0.00018811002988575382, + "loss": 2.4445, "step": 621 }, { - "epoch": 0.19919935948759007, - "grad_norm": 0.5234375, - "learning_rate": 4.85219245110494e-05, - "loss": 2.5425, + "epoch": 0.24166484543843034, + "grad_norm": 0.87109375, + "learning_rate": 0.00018804576943500333, + "loss": 2.4991, "step": 622 }, { - "epoch": 0.19951961569255405, - "grad_norm": 0.373046875, - "learning_rate": 4.8512438374396095e-05, - "loss": 2.3095, + "epoch": 0.24205337412884584, + "grad_norm": 0.984375, + "learning_rate": 0.00018798134683827464, + "loss": 2.4199, "step": 623 }, { - "epoch": 0.199839871897518, - "grad_norm": 0.87109375, - "learning_rate": 4.850292282781949e-05, - "loss": 2.5806, + "epoch": 0.2424419028192613, + "grad_norm": 0.76953125, + "learning_rate": 0.000187916762214209, + "loss": 2.3854, "step": 624 }, { - "epoch": 0.200160128102482, - "grad_norm": 0.609375, - "learning_rate": 4.849337788322185e-05, - "loss": 2.4287, + "epoch": 0.2428304315096768, + "grad_norm": 1.0, + "learning_rate": 0.00018785201568174617, + "loss": 2.4612, "step": 625 }, { - "epoch": 0.20048038430744597, - "grad_norm": 0.55078125, - "learning_rate": 4.848380355254222e-05, - "loss": 2.6023, + "epoch": 0.2432189602000923, + "grad_norm": 0.9296875, + "learning_rate": 0.00018778710736012396, + "loss": 2.4219, "step": 626 }, { - "epoch": 0.20080064051240992, - "grad_norm": 0.625, - "learning_rate": 4.847419984775641e-05, - "loss": 2.0865, + "epoch": 0.24360748889050776, + "grad_norm": 0.828125, + "learning_rate": 0.0001877220373688782, + "loss": 2.4123, "step": 627 }, { - "epoch": 0.2011208967173739, - "grad_norm": 1.6484375, - "learning_rate": 4.8464566780876954e-05, - "loss": 1.8258, + "epoch": 0.24399601758092324, + "grad_norm": 0.83984375, + "learning_rate": 0.0001876568058278425, + "loss": 2.4361, "step": 628 }, { - "epoch": 0.20144115292233786, - "grad_norm": 0.515625, - "learning_rate": 4.845490436395314e-05, - "loss": 2.2667, + "epoch": 0.24438454627133874, + "grad_norm": 0.828125, + "learning_rate": 0.00018759141285714783, + "loss": 2.4355, "step": 629 }, { - "epoch": 0.20176140912730184, - "grad_norm": 0.5234375, - "learning_rate": 4.8445212609070934e-05, - "loss": 2.394, + "epoch": 0.2447730749617542, + "grad_norm": 0.8515625, + "learning_rate": 0.00018752585857722266, + "loss": 2.4675, "step": 630 }, { - "epoch": 0.20208166533226582, - "grad_norm": 0.58984375, - "learning_rate": 4.8435491528353026e-05, - "loss": 2.6123, + "epoch": 0.24516160365216969, + "grad_norm": 0.85546875, + "learning_rate": 0.00018746014310879228, + "loss": 2.4521, "step": 631 }, { - "epoch": 0.20240192153722977, - "grad_norm": 1.25, - "learning_rate": 4.8425741133958774e-05, - "loss": 2.2056, + "epoch": 0.24555013234258516, + "grad_norm": 0.82421875, + "learning_rate": 0.00018739426657287907, + "loss": 2.435, "step": 632 }, { - "epoch": 0.20272217774219375, - "grad_norm": 0.4453125, - "learning_rate": 4.841596143808421e-05, - "loss": 2.6455, + "epoch": 0.24593866103300066, + "grad_norm": 0.8359375, + "learning_rate": 0.0001873282290908019, + "loss": 2.4109, "step": 633 }, { - "epoch": 0.20304243394715774, - "grad_norm": 0.37890625, - "learning_rate": 4.840615245296201e-05, - "loss": 2.1008, + "epoch": 0.24632718972341613, + "grad_norm": 0.79296875, + "learning_rate": 0.00018726203078417604, + "loss": 2.4263, "step": 634 }, { - "epoch": 0.2033626901521217, - "grad_norm": 0.453125, - "learning_rate": 4.8396314190861495e-05, - "loss": 1.9474, + "epoch": 0.2467157184138316, + "grad_norm": 0.80078125, + "learning_rate": 0.00018719567177491296, + "loss": 2.3932, "step": 635 }, { - "epoch": 0.20368294635708567, - "grad_norm": 0.578125, - "learning_rate": 4.838644666408858e-05, - "loss": 2.1314, + "epoch": 0.2471042471042471, + "grad_norm": 0.80078125, + "learning_rate": 0.0001871291521852201, + "loss": 2.3958, "step": 636 }, { - "epoch": 0.20400320256204965, - "grad_norm": 0.4375, - "learning_rate": 4.8376549884985825e-05, - "loss": 2.4542, + "epoch": 0.24749277579466258, + "grad_norm": 0.9765625, + "learning_rate": 0.0001870624721376006, + "loss": 2.4574, "step": 637 }, { - "epoch": 0.2043234587670136, - "grad_norm": 0.40234375, - "learning_rate": 4.836662386593235e-05, - "loss": 2.2233, + "epoch": 0.24788130448507806, + "grad_norm": 0.83984375, + "learning_rate": 0.0001869956317548531, + "loss": 2.4139, "step": 638 }, { - "epoch": 0.20464371497197759, - "grad_norm": 0.4296875, - "learning_rate": 4.835666861934387e-05, - "loss": 1.8406, + "epoch": 0.24826983317549356, + "grad_norm": 0.8984375, + "learning_rate": 0.00018692863116007152, + "loss": 2.4308, "step": 639 }, { - "epoch": 0.20496397117694154, - "grad_norm": 0.458984375, - "learning_rate": 4.834668415767263e-05, - "loss": 2.3927, + "epoch": 0.24865836186590903, + "grad_norm": 0.78125, + "learning_rate": 0.0001868614704766449, + "loss": 2.3843, "step": 640 }, { - "epoch": 0.20528422738190552, - "grad_norm": 0.458984375, - "learning_rate": 4.833667049340744e-05, - "loss": 2.4086, + "epoch": 0.2490468905563245, + "grad_norm": 0.84765625, + "learning_rate": 0.00018679414982825693, + "loss": 2.4272, "step": 641 }, { - "epoch": 0.2056044835868695, - "grad_norm": 0.53125, - "learning_rate": 4.832662763907363e-05, - "loss": 1.7985, + "epoch": 0.24943541924674, + "grad_norm": 0.86328125, + "learning_rate": 0.00018672666933888603, + "loss": 2.4451, "step": 642 }, { - "epoch": 0.20592473979183346, - "grad_norm": 0.423828125, - "learning_rate": 4.831655560723306e-05, - "loss": 2.2216, + "epoch": 0.24982394793715548, + "grad_norm": 0.92578125, + "learning_rate": 0.00018665902913280496, + "loss": 2.4515, "step": 643 }, { - "epoch": 0.20624499599679744, - "grad_norm": 1.3515625, - "learning_rate": 4.8306454410484055e-05, - "loss": 2.4505, + "epoch": 0.25021247662757096, + "grad_norm": 0.85546875, + "learning_rate": 0.00018659122933458065, + "loss": 2.399, "step": 644 }, { - "epoch": 0.20656525220176142, - "grad_norm": 0.60546875, - "learning_rate": 4.829632406146143e-05, - "loss": 2.1029, + "epoch": 0.25060100531798646, + "grad_norm": 0.796875, + "learning_rate": 0.00018652327006907386, + "loss": 2.5024, "step": 645 }, { - "epoch": 0.20688550840672537, - "grad_norm": 0.5, - "learning_rate": 4.828616457283648e-05, - "loss": 2.4966, + "epoch": 0.25098953400840196, + "grad_norm": 0.83984375, + "learning_rate": 0.00018645515146143904, + "loss": 2.3793, "step": 646 }, { - "epoch": 0.20720576461168935, - "grad_norm": 0.384765625, - "learning_rate": 4.8275975957316943e-05, - "loss": 1.7865, + "epoch": 0.2513780626988174, + "grad_norm": 0.80859375, + "learning_rate": 0.0001863868736371241, + "loss": 2.4433, "step": 647 }, { - "epoch": 0.20752602081665333, - "grad_norm": 0.4140625, - "learning_rate": 4.826575822764696e-05, - "loss": 1.674, + "epoch": 0.2517665913892329, + "grad_norm": 0.88671875, + "learning_rate": 0.00018631843672187022, + "loss": 2.4498, "step": 648 }, { - "epoch": 0.2078462770216173, - "grad_norm": 0.404296875, - "learning_rate": 4.825551139660715e-05, - "loss": 1.9932, + "epoch": 0.25215512007964835, + "grad_norm": 0.859375, + "learning_rate": 0.00018624984084171148, + "loss": 2.4227, "step": 649 }, { - "epoch": 0.20816653322658127, - "grad_norm": 0.84765625, - "learning_rate": 4.824523547701445e-05, - "loss": 2.18, + "epoch": 0.25254364877006386, + "grad_norm": 0.83203125, + "learning_rate": 0.00018618108612297474, + "loss": 2.5026, "step": 650 }, { - "epoch": 0.20848678943154525, - "grad_norm": 1.7421875, - "learning_rate": 4.8234930481722264e-05, - "loss": 1.9729, + "epoch": 0.25293217746047936, + "grad_norm": 0.86328125, + "learning_rate": 0.0001861121726922794, + "loss": 2.4045, "step": 651 }, { - "epoch": 0.2088070456365092, - "grad_norm": 0.5625, - "learning_rate": 4.8224596423620315e-05, - "loss": 2.0876, + "epoch": 0.2533207061508948, + "grad_norm": 0.8671875, + "learning_rate": 0.00018604310067653712, + "loss": 2.445, "step": 652 }, { - "epoch": 0.20912730184147318, - "grad_norm": 0.8359375, - "learning_rate": 4.821423331563469e-05, - "loss": 2.0135, + "epoch": 0.2537092348413103, + "grad_norm": 0.8203125, + "learning_rate": 0.0001859738702029516, + "loss": 2.3895, "step": 653 }, { - "epoch": 0.20944755804643714, - "grad_norm": 0.416015625, - "learning_rate": 4.820384117072781e-05, - "loss": 2.3225, + "epoch": 0.2540977635317258, + "grad_norm": 0.88671875, + "learning_rate": 0.00018590448139901847, + "loss": 2.4448, "step": 654 }, { - "epoch": 0.20976781425140112, - "grad_norm": 1.65625, - "learning_rate": 4.819342000189844e-05, - "loss": 1.6923, + "epoch": 0.25448629222214125, + "grad_norm": 0.94140625, + "learning_rate": 0.00018583493439252476, + "loss": 2.4411, "step": 655 }, { - "epoch": 0.2100880704563651, - "grad_norm": 0.435546875, - "learning_rate": 4.818296982218161e-05, - "loss": 2.1741, + "epoch": 0.25487482091255675, + "grad_norm": 0.83984375, + "learning_rate": 0.00018576522931154895, + "loss": 2.4646, "step": 656 }, { - "epoch": 0.21040832666132905, - "grad_norm": 0.671875, - "learning_rate": 4.817249064464865e-05, - "loss": 2.285, + "epoch": 0.25526334960297226, + "grad_norm": 0.8359375, + "learning_rate": 0.00018569536628446065, + "loss": 2.4226, "step": 657 }, { - "epoch": 0.21072858286629303, - "grad_norm": 0.6484375, - "learning_rate": 4.816198248240718e-05, - "loss": 2.1315, + "epoch": 0.2556518782933877, + "grad_norm": 0.875, + "learning_rate": 0.0001856253454399203, + "loss": 2.4317, "step": 658 }, { - "epoch": 0.21104883907125702, - "grad_norm": 0.498046875, - "learning_rate": 4.815144534860106e-05, - "loss": 1.9737, + "epoch": 0.2560404069838032, + "grad_norm": 0.93359375, + "learning_rate": 0.000185555166906879, + "loss": 2.4337, "step": 659 }, { - "epoch": 0.21136909527622097, - "grad_norm": 0.453125, - "learning_rate": 4.81408792564104e-05, - "loss": 1.6167, + "epoch": 0.2564289356742187, + "grad_norm": 0.83203125, + "learning_rate": 0.00018548483081457827, + "loss": 2.4088, "step": 660 }, { - "epoch": 0.21168935148118495, - "grad_norm": 0.421875, - "learning_rate": 4.8130284219051494e-05, - "loss": 2.4866, + "epoch": 0.25681746436463415, + "grad_norm": 0.890625, + "learning_rate": 0.00018541433729254972, + "loss": 2.4964, "step": 661 }, { - "epoch": 0.21200960768614893, - "grad_norm": 0.7578125, - "learning_rate": 4.81196602497769e-05, - "loss": 2.2202, + "epoch": 0.25720599305504965, + "grad_norm": 0.93359375, + "learning_rate": 0.00018534368647061495, + "loss": 2.4715, "step": 662 }, { - "epoch": 0.21232986389111289, - "grad_norm": 0.51953125, - "learning_rate": 4.810900736187531e-05, - "loss": 2.1335, + "epoch": 0.25759452174546515, + "grad_norm": 0.8671875, + "learning_rate": 0.00018527287847888523, + "loss": 2.4574, "step": 663 }, { - "epoch": 0.21265012009607687, - "grad_norm": 0.58203125, - "learning_rate": 4.809832556867164e-05, - "loss": 2.0997, + "epoch": 0.2579830504358806, + "grad_norm": 0.80078125, + "learning_rate": 0.0001852019134477613, + "loss": 2.4681, "step": 664 }, { - "epoch": 0.21297037630104082, - "grad_norm": 1.9765625, - "learning_rate": 4.80876148835269e-05, - "loss": 2.5337, + "epoch": 0.2583715791262961, + "grad_norm": 0.9609375, + "learning_rate": 0.00018513079150793305, + "loss": 2.4757, "step": 665 }, { - "epoch": 0.2132906325060048, - "grad_norm": 0.5390625, - "learning_rate": 4.80768753198383e-05, - "loss": 2.4361, + "epoch": 0.2587601078167116, + "grad_norm": 0.83984375, + "learning_rate": 0.00018505951279037934, + "loss": 2.4098, "step": 666 }, { - "epoch": 0.21361088871096878, - "grad_norm": 0.408203125, - "learning_rate": 4.8066106891039135e-05, - "loss": 2.7403, + "epoch": 0.25914863650712705, + "grad_norm": 0.76171875, + "learning_rate": 0.0001849880774263678, + "loss": 2.353, "step": 667 }, { - "epoch": 0.21393114491593274, - "grad_norm": 0.44921875, - "learning_rate": 4.805530961059881e-05, - "loss": 1.8824, + "epoch": 0.25953716519754255, + "grad_norm": 0.890625, + "learning_rate": 0.00018491648554745457, + "loss": 2.4913, "step": 668 }, { - "epoch": 0.21425140112089672, - "grad_norm": 0.51953125, - "learning_rate": 4.804448349202283e-05, - "loss": 1.9837, + "epoch": 0.25992569388795805, + "grad_norm": 0.796875, + "learning_rate": 0.0001848447372854839, + "loss": 2.4221, "step": 669 }, { - "epoch": 0.2145716573258607, - "grad_norm": 0.52734375, - "learning_rate": 4.803362854885276e-05, - "loss": 2.0329, + "epoch": 0.2603142225783735, + "grad_norm": 0.81640625, + "learning_rate": 0.00018477283277258812, + "loss": 2.3949, "step": 670 }, { - "epoch": 0.21489191353082465, - "grad_norm": 0.91015625, - "learning_rate": 4.802274479466623e-05, - "loss": 2.314, + "epoch": 0.260702751268789, + "grad_norm": 0.80859375, + "learning_rate": 0.00018470077214118732, + "loss": 2.4428, "step": 671 }, { - "epoch": 0.21521216973578863, - "grad_norm": 0.44140625, - "learning_rate": 4.8011832243076895e-05, - "loss": 2.324, + "epoch": 0.2610912799592045, + "grad_norm": 0.75390625, + "learning_rate": 0.0001846285555239891, + "loss": 2.4141, "step": 672 }, { - "epoch": 0.21553242594075261, - "grad_norm": 0.53515625, - "learning_rate": 4.8000890907734434e-05, - "loss": 1.7424, + "epoch": 0.26147980864961995, + "grad_norm": 0.90625, + "learning_rate": 0.00018455618305398836, + "loss": 2.3777, "step": 673 }, { - "epoch": 0.21585268214571657, - "grad_norm": 0.37109375, - "learning_rate": 4.7989920802324537e-05, - "loss": 1.981, + "epoch": 0.26186833734003545, + "grad_norm": 0.84375, + "learning_rate": 0.00018448365486446687, + "loss": 2.412, "step": 674 }, { - "epoch": 0.21617293835068055, - "grad_norm": 0.412109375, - "learning_rate": 4.797892194056889e-05, - "loss": 1.6027, + "epoch": 0.26225686603045095, + "grad_norm": 0.828125, + "learning_rate": 0.00018441097108899332, + "loss": 2.4766, "step": 675 }, { - "epoch": 0.2164931945556445, - "grad_norm": 0.48828125, - "learning_rate": 4.79678943362251e-05, - "loss": 2.5835, + "epoch": 0.2626453947208664, + "grad_norm": 0.8515625, + "learning_rate": 0.0001843381318614229, + "loss": 2.4765, "step": 676 }, { - "epoch": 0.21681345076060848, - "grad_norm": 0.7734375, - "learning_rate": 4.79568380030868e-05, - "loss": 2.4851, + "epoch": 0.2630339234112819, + "grad_norm": 0.9140625, + "learning_rate": 0.0001842651373158971, + "loss": 2.4458, "step": 677 }, { - "epoch": 0.21713370696557246, - "grad_norm": 0.4296875, - "learning_rate": 4.794575295498348e-05, - "loss": 2.6476, + "epoch": 0.2634224521016974, + "grad_norm": 0.83984375, + "learning_rate": 0.0001841919875868433, + "loss": 2.3816, "step": 678 }, { - "epoch": 0.21745396317053642, - "grad_norm": 0.427734375, - "learning_rate": 4.793463920578061e-05, - "loss": 2.6871, + "epoch": 0.26381098079211285, + "grad_norm": 0.8984375, + "learning_rate": 0.00018411868280897488, + "loss": 2.4251, "step": 679 }, { - "epoch": 0.2177742193755004, - "grad_norm": 0.279296875, - "learning_rate": 4.792349676937953e-05, - "loss": 2.124, + "epoch": 0.26419950948252835, + "grad_norm": 0.83984375, + "learning_rate": 0.00018404522311729062, + "loss": 2.4098, "step": 680 }, { - "epoch": 0.21809447558046438, - "grad_norm": 0.546875, - "learning_rate": 4.791232565971745e-05, - "loss": 2.3615, + "epoch": 0.26458803817294385, + "grad_norm": 0.828125, + "learning_rate": 0.00018397160864707467, + "loss": 2.4377, "step": 681 }, { - "epoch": 0.21841473178542833, - "grad_norm": 0.462890625, - "learning_rate": 4.7901125890767484e-05, - "loss": 1.8175, + "epoch": 0.2649765668633593, + "grad_norm": 0.91796875, + "learning_rate": 0.00018389783953389613, + "loss": 2.4096, "step": 682 }, { - "epoch": 0.21873498799039232, - "grad_norm": 0.75390625, - "learning_rate": 4.788989747653857e-05, - "loss": 2.2353, + "epoch": 0.2653650955537748, + "grad_norm": 0.81640625, + "learning_rate": 0.00018382391591360896, + "loss": 2.4866, "step": 683 }, { - "epoch": 0.2190552441953563, - "grad_norm": 0.52734375, - "learning_rate": 4.787864043107546e-05, - "loss": 1.9429, + "epoch": 0.2657536242441903, + "grad_norm": 0.8203125, + "learning_rate": 0.00018374983792235163, + "loss": 2.4486, "step": 684 }, { - "epoch": 0.21937550040032025, - "grad_norm": 0.48828125, - "learning_rate": 4.786735476845876e-05, - "loss": 2.1238, + "epoch": 0.26614215293460575, + "grad_norm": 0.78515625, + "learning_rate": 0.00018367560569654693, + "loss": 2.4155, "step": 685 }, { - "epoch": 0.21969575660528423, - "grad_norm": 0.361328125, - "learning_rate": 4.7856040502804836e-05, - "loss": 1.7298, + "epoch": 0.26653068162502125, + "grad_norm": 0.7734375, + "learning_rate": 0.0001836012193729017, + "loss": 2.3801, "step": 686 }, { - "epoch": 0.22001601281024818, - "grad_norm": 0.51953125, - "learning_rate": 4.784469764826585e-05, - "loss": 1.9109, + "epoch": 0.26691921031543675, + "grad_norm": 0.77734375, + "learning_rate": 0.00018352667908840654, + "loss": 2.3832, "step": 687 }, { - "epoch": 0.22033626901521217, - "grad_norm": 0.474609375, - "learning_rate": 4.783332621902971e-05, - "loss": 2.7153, + "epoch": 0.2673077390058522, + "grad_norm": 0.80859375, + "learning_rate": 0.00018345198498033558, + "loss": 2.4371, "step": 688 }, { - "epoch": 0.22065652522017615, - "grad_norm": 0.6171875, - "learning_rate": 4.782192622932008e-05, - "loss": 2.3001, + "epoch": 0.2676962676962677, + "grad_norm": 0.796875, + "learning_rate": 0.00018337713718624623, + "loss": 2.5273, "step": 689 }, { - "epoch": 0.2209767814251401, - "grad_norm": 0.6640625, - "learning_rate": 4.7810497693396327e-05, - "loss": 2.0766, + "epoch": 0.2680847963866832, + "grad_norm": 0.8046875, + "learning_rate": 0.00018330213584397896, + "loss": 2.4359, "step": 690 }, { - "epoch": 0.22129703763010408, - "grad_norm": 0.640625, - "learning_rate": 4.779904062555356e-05, - "loss": 2.2382, + "epoch": 0.26847332507709865, + "grad_norm": 0.76953125, + "learning_rate": 0.000183226981091657, + "loss": 2.5058, "step": 691 }, { - "epoch": 0.22161729383506806, - "grad_norm": 0.51171875, - "learning_rate": 4.7787555040122534e-05, - "loss": 2.6912, + "epoch": 0.26886185376751415, + "grad_norm": 0.80078125, + "learning_rate": 0.0001831516730676861, + "loss": 2.4414, "step": 692 }, { - "epoch": 0.22193755004003202, - "grad_norm": 0.5703125, - "learning_rate": 4.777604095146971e-05, - "loss": 2.3497, + "epoch": 0.26925038245792965, + "grad_norm": 0.84765625, + "learning_rate": 0.00018307621191075425, + "loss": 2.4316, "step": 693 }, { - "epoch": 0.222257806244996, - "grad_norm": 0.458984375, - "learning_rate": 4.7764498373997194e-05, - "loss": 2.4798, + "epoch": 0.2696389111483451, + "grad_norm": 0.8046875, + "learning_rate": 0.00018300059775983152, + "loss": 2.4051, "step": 694 }, { - "epoch": 0.22257806244995998, - "grad_norm": 0.89453125, - "learning_rate": 4.775292732214269e-05, - "loss": 2.6314, + "epoch": 0.2700274398387606, + "grad_norm": 0.828125, + "learning_rate": 0.0001829248307541697, + "loss": 2.4377, "step": 695 }, { - "epoch": 0.22289831865492393, - "grad_norm": 0.51953125, - "learning_rate": 4.774132781037958e-05, - "loss": 2.2279, + "epoch": 0.2704159685291761, + "grad_norm": 0.78125, + "learning_rate": 0.000182848911033302, + "loss": 2.3461, "step": 696 }, { - "epoch": 0.2232185748598879, - "grad_norm": 0.53515625, - "learning_rate": 4.7729699853216784e-05, - "loss": 2.6009, + "epoch": 0.27080449721959154, + "grad_norm": 0.88671875, + "learning_rate": 0.000182772838737043, + "loss": 2.4788, "step": 697 }, { - "epoch": 0.2235388310648519, - "grad_norm": 0.546875, - "learning_rate": 4.771804346519886e-05, - "loss": 2.1287, + "epoch": 0.27119302591000705, + "grad_norm": 0.78515625, + "learning_rate": 0.00018269661400548823, + "loss": 2.4301, "step": 698 }, { - "epoch": 0.22385908726981585, - "grad_norm": 0.392578125, - "learning_rate": 4.770635866090587e-05, - "loss": 2.0747, + "epoch": 0.27158155460042255, + "grad_norm": 0.8984375, + "learning_rate": 0.0001826202369790139, + "loss": 2.2877, "step": 699 }, { - "epoch": 0.22417934347477983, - "grad_norm": 1.2578125, - "learning_rate": 4.769464545495347e-05, - "loss": 2.1692, + "epoch": 0.271970083290838, + "grad_norm": 0.91796875, + "learning_rate": 0.00018254370779827668, + "loss": 2.4765, "step": 700 }, { - "epoch": 0.22449959967974378, - "grad_norm": 0.5625, - "learning_rate": 4.768290386199281e-05, - "loss": 2.1745, + "epoch": 0.2723586119812535, + "grad_norm": 0.79296875, + "learning_rate": 0.00018246702660421356, + "loss": 2.371, "step": 701 }, { - "epoch": 0.22481985588470776, - "grad_norm": 1.0234375, - "learning_rate": 4.767113389671055e-05, - "loss": 2.4571, + "epoch": 0.272747140671669, + "grad_norm": 0.86328125, + "learning_rate": 0.00018239019353804133, + "loss": 2.4542, "step": 702 }, { - "epoch": 0.22514011208967175, - "grad_norm": 0.486328125, - "learning_rate": 4.7659335573828855e-05, - "loss": 2.1443, + "epoch": 0.27313566936208444, + "grad_norm": 1.2265625, + "learning_rate": 0.00018231320874125654, + "loss": 2.4434, "step": 703 }, { - "epoch": 0.2254603682946357, - "grad_norm": 0.365234375, - "learning_rate": 4.7647508908105354e-05, - "loss": 2.645, + "epoch": 0.27352419805249994, + "grad_norm": 1.890625, + "learning_rate": 0.00018223607235563524, + "loss": 2.4492, "step": 704 }, { - "epoch": 0.22578062449959968, - "grad_norm": 0.50390625, - "learning_rate": 4.7635653914333123e-05, - "loss": 2.5967, + "epoch": 0.27391272674291545, + "grad_norm": 0.91015625, + "learning_rate": 0.0001821587845232325, + "loss": 2.4152, "step": 705 }, { - "epoch": 0.22610088070456366, - "grad_norm": 0.3828125, - "learning_rate": 4.7623770607340656e-05, - "loss": 1.8238, + "epoch": 0.2743012554333309, + "grad_norm": 1.0625, + "learning_rate": 0.00018208134538638245, + "loss": 2.4712, "step": 706 }, { - "epoch": 0.22642113690952761, - "grad_norm": 0.4765625, - "learning_rate": 4.761185900199191e-05, - "loss": 2.1985, + "epoch": 0.2746897841237464, + "grad_norm": 0.8046875, + "learning_rate": 0.00018200375508769772, + "loss": 2.4033, "step": 707 }, { - "epoch": 0.2267413931144916, - "grad_norm": 1.15625, - "learning_rate": 4.7599919113186184e-05, - "loss": 1.9056, + "epoch": 0.2750783128141619, + "grad_norm": 0.84765625, + "learning_rate": 0.0001819260137700694, + "loss": 2.4302, "step": 708 }, { - "epoch": 0.22706164931945558, - "grad_norm": 0.53515625, - "learning_rate": 4.7587950955858184e-05, - "loss": 2.0725, + "epoch": 0.27546684150457734, + "grad_norm": 0.93359375, + "learning_rate": 0.00018184812157666667, + "loss": 2.3352, "step": 709 }, { - "epoch": 0.22738190552441953, - "grad_norm": 0.62109375, - "learning_rate": 4.757595454497798e-05, - "loss": 2.1442, + "epoch": 0.27585537019499284, + "grad_norm": 0.8984375, + "learning_rate": 0.00018177007865093664, + "loss": 2.4609, "step": 710 }, { - "epoch": 0.2277021617293835, - "grad_norm": 0.34375, - "learning_rate": 4.756392989555094e-05, - "loss": 1.8137, + "epoch": 0.27624389888540835, + "grad_norm": 0.765625, + "learning_rate": 0.0001816918851366039, + "loss": 2.4525, "step": 711 }, { - "epoch": 0.22802241793434747, - "grad_norm": 0.515625, - "learning_rate": 4.755187702261782e-05, - "loss": 2.4148, + "epoch": 0.2766324275758238, + "grad_norm": 0.9296875, + "learning_rate": 0.00018161354117767045, + "loss": 2.4259, "step": 712 }, { - "epoch": 0.22834267413931145, - "grad_norm": 0.54296875, - "learning_rate": 4.753979594125463e-05, - "loss": 2.579, + "epoch": 0.2770209562662393, + "grad_norm": 0.8515625, + "learning_rate": 0.00018153504691841528, + "loss": 2.4029, "step": 713 }, { - "epoch": 0.22866293034427543, - "grad_norm": 0.55078125, - "learning_rate": 4.752768666657267e-05, - "loss": 2.4936, + "epoch": 0.2774094849566548, + "grad_norm": 0.80078125, + "learning_rate": 0.00018145640250339425, + "loss": 2.3984, "step": 714 }, { - "epoch": 0.22898318654923938, - "grad_norm": 0.498046875, - "learning_rate": 4.751554921371852e-05, - "loss": 2.104, + "epoch": 0.27779801364707024, + "grad_norm": 0.83984375, + "learning_rate": 0.00018137760807743965, + "loss": 2.4384, "step": 715 }, { - "epoch": 0.22930344275420336, - "grad_norm": 0.3984375, - "learning_rate": 4.7503383597873994e-05, - "loss": 2.0582, + "epoch": 0.27818654233748574, + "grad_norm": 0.8671875, + "learning_rate": 0.00018129866378566013, + "loss": 2.427, "step": 716 }, { - "epoch": 0.22962369895916734, - "grad_norm": 0.435546875, - "learning_rate": 4.7491189834256147e-05, - "loss": 2.2863, + "epoch": 0.27857507102790124, + "grad_norm": 0.87109375, + "learning_rate": 0.00018121956977344033, + "loss": 2.4698, "step": 717 }, { - "epoch": 0.2299439551641313, - "grad_norm": 0.5703125, - "learning_rate": 4.747896793811721e-05, - "loss": 2.4744, + "epoch": 0.2789635997183167, + "grad_norm": 0.81640625, + "learning_rate": 0.00018114032618644053, + "loss": 2.3839, "step": 718 }, { - "epoch": 0.23026421136909528, - "grad_norm": 0.4921875, - "learning_rate": 4.746671792474465e-05, - "loss": 2.4798, + "epoch": 0.2793521284087322, + "grad_norm": 0.8203125, + "learning_rate": 0.0001810609331705965, + "loss": 2.4457, "step": 719 }, { - "epoch": 0.23058446757405926, - "grad_norm": 0.60546875, - "learning_rate": 4.7454439809461064e-05, - "loss": 2.4989, + "epoch": 0.27974065709914764, + "grad_norm": 1.0, + "learning_rate": 0.00018098139087211927, + "loss": 2.4193, "step": 720 }, { - "epoch": 0.2309047237790232, - "grad_norm": 0.482421875, - "learning_rate": 4.7442133607624216e-05, - "loss": 2.4011, + "epoch": 0.28012918578956314, + "grad_norm": 1.09375, + "learning_rate": 0.00018090169943749476, + "loss": 2.4061, "step": 721 }, { - "epoch": 0.2312249799839872, - "grad_norm": 0.498046875, - "learning_rate": 4.7429799334627e-05, - "loss": 1.8769, + "epoch": 0.28051771447997864, + "grad_norm": 0.91796875, + "learning_rate": 0.00018082185901348346, + "loss": 2.3813, "step": 722 }, { - "epoch": 0.23154523618895115, - "grad_norm": 0.4296875, - "learning_rate": 4.7417437005897415e-05, - "loss": 2.1301, + "epoch": 0.2809062431703941, + "grad_norm": 0.94140625, + "learning_rate": 0.00018074186974712032, + "loss": 2.4133, "step": 723 }, { - "epoch": 0.23186549239391513, - "grad_norm": 0.484375, - "learning_rate": 4.740504663689857e-05, - "loss": 1.8229, + "epoch": 0.2812947718608096, + "grad_norm": 0.94921875, + "learning_rate": 0.0001806617317857144, + "loss": 2.4118, "step": 724 }, { - "epoch": 0.2321857485988791, - "grad_norm": 0.48828125, - "learning_rate": 4.739262824312863e-05, - "loss": 2.2332, + "epoch": 0.2816833005512251, + "grad_norm": 0.80859375, + "learning_rate": 0.00018058144527684854, + "loss": 2.4778, "step": 725 }, { - "epoch": 0.23250600480384306, - "grad_norm": 0.462890625, - "learning_rate": 4.7380181840120825e-05, - "loss": 2.4415, + "epoch": 0.28207182924164054, + "grad_norm": 1.09375, + "learning_rate": 0.00018050101036837926, + "loss": 2.4505, "step": 726 }, { - "epoch": 0.23282626100880705, - "grad_norm": 0.80078125, - "learning_rate": 4.736770744344341e-05, - "loss": 2.4059, + "epoch": 0.28246035793205604, + "grad_norm": 1.078125, + "learning_rate": 0.00018042042720843623, + "loss": 2.3984, "step": 727 }, { - "epoch": 0.23314651721377103, - "grad_norm": 0.58203125, - "learning_rate": 4.735520506869966e-05, - "loss": 2.3278, + "epoch": 0.28284888662247154, + "grad_norm": 0.8828125, + "learning_rate": 0.0001803396959454222, + "loss": 2.4154, "step": 728 }, { - "epoch": 0.23346677341873498, - "grad_norm": 0.439453125, - "learning_rate": 4.734267473152785e-05, - "loss": 2.3155, + "epoch": 0.283237415312887, + "grad_norm": 1.078125, + "learning_rate": 0.00018025881672801273, + "loss": 2.4028, "step": 729 }, { - "epoch": 0.23378702962369896, - "grad_norm": 0.40234375, - "learning_rate": 4.7330116447601225e-05, - "loss": 1.8021, + "epoch": 0.2836259440033025, + "grad_norm": 0.9140625, + "learning_rate": 0.0001801777897051558, + "loss": 2.4469, "step": 730 }, { - "epoch": 0.23410728582866294, - "grad_norm": 0.4765625, - "learning_rate": 4.731753023262798e-05, - "loss": 2.0426, + "epoch": 0.284014472693718, + "grad_norm": 0.9609375, + "learning_rate": 0.00018009661502607158, + "loss": 2.2968, "step": 731 }, { - "epoch": 0.2344275420336269, - "grad_norm": 0.59765625, - "learning_rate": 4.730491610235128e-05, - "loss": 1.5206, + "epoch": 0.28440300138413344, + "grad_norm": 1.03125, + "learning_rate": 0.0001800152928402522, + "loss": 2.4668, "step": 732 }, { - "epoch": 0.23474779823859088, - "grad_norm": 0.5859375, - "learning_rate": 4.729227407254916e-05, - "loss": 2.1711, + "epoch": 0.28479153007454894, + "grad_norm": 1.0546875, + "learning_rate": 0.00017993382329746133, + "loss": 2.4243, "step": 733 }, { - "epoch": 0.23506805444355483, - "grad_norm": 0.578125, - "learning_rate": 4.727960415903458e-05, - "loss": 2.246, + "epoch": 0.28518005876496444, + "grad_norm": 0.7265625, + "learning_rate": 0.0001798522065477342, + "loss": 2.3054, "step": 734 }, { - "epoch": 0.2353883106485188, - "grad_norm": 0.44140625, - "learning_rate": 4.7266906377655375e-05, - "loss": 1.8435, + "epoch": 0.2855685874553799, + "grad_norm": 0.96875, + "learning_rate": 0.00017977044274137703, + "loss": 2.3318, "step": 735 }, { - "epoch": 0.2357085668534828, - "grad_norm": 0.423828125, - "learning_rate": 4.725418074429423e-05, - "loss": 1.9849, + "epoch": 0.2859571161457954, + "grad_norm": 0.96875, + "learning_rate": 0.00017968853202896684, + "loss": 2.3392, "step": 736 }, { - "epoch": 0.23602882305844675, - "grad_norm": 0.41015625, - "learning_rate": 4.724142727486869e-05, - "loss": 2.6527, + "epoch": 0.2863456448362109, + "grad_norm": 0.796875, + "learning_rate": 0.00017960647456135125, + "loss": 2.3525, "step": 737 }, { - "epoch": 0.23634907926341073, - "grad_norm": 0.412109375, - "learning_rate": 4.722864598533108e-05, - "loss": 2.5422, + "epoch": 0.28673417352662633, + "grad_norm": 0.92578125, + "learning_rate": 0.0001795242704896481, + "loss": 2.4782, "step": 738 }, { - "epoch": 0.2366693354683747, - "grad_norm": 0.60546875, - "learning_rate": 4.721583689166855e-05, - "loss": 2.0435, + "epoch": 0.28712270221704184, + "grad_norm": 0.8828125, + "learning_rate": 0.00017944191996524525, + "loss": 2.4221, "step": 739 }, { - "epoch": 0.23698959167333866, - "grad_norm": 0.8515625, - "learning_rate": 4.720300000990302e-05, - "loss": 1.571, + "epoch": 0.28751123090745734, + "grad_norm": 0.86328125, + "learning_rate": 0.00017935942313980022, + "loss": 2.4797, "step": 740 }, { - "epoch": 0.23730984787830264, - "grad_norm": 0.51953125, - "learning_rate": 4.719013535609119e-05, - "loss": 1.9472, + "epoch": 0.2878997595978728, + "grad_norm": 0.79296875, + "learning_rate": 0.00017927678016524005, + "loss": 2.4934, "step": 741 }, { - "epoch": 0.23763010408326662, - "grad_norm": 0.5078125, - "learning_rate": 4.717724294632446e-05, - "loss": 2.4293, + "epoch": 0.2882882882882883, + "grad_norm": 0.8984375, + "learning_rate": 0.00017919399119376087, + "loss": 2.4664, "step": 742 }, { - "epoch": 0.23795036028823058, - "grad_norm": 0.5, - "learning_rate": 4.716432279672897e-05, - "loss": 2.4161, + "epoch": 0.2886768169787038, + "grad_norm": 1.140625, + "learning_rate": 0.00017911105637782767, + "loss": 2.4354, "step": 743 }, { - "epoch": 0.23827061649319456, - "grad_norm": 0.486328125, - "learning_rate": 4.7151374923465554e-05, - "loss": 2.4954, + "epoch": 0.28906534566911923, + "grad_norm": 0.8125, + "learning_rate": 0.00017902797587017406, + "loss": 2.4726, "step": 744 }, { - "epoch": 0.23859087269815854, - "grad_norm": 1.1953125, - "learning_rate": 4.7138399342729723e-05, - "loss": 2.3457, + "epoch": 0.28945387435953474, + "grad_norm": 1.0, + "learning_rate": 0.00017894474982380193, + "loss": 2.4309, "step": 745 }, { - "epoch": 0.2389111289031225, - "grad_norm": 1.6015625, - "learning_rate": 4.712539607075164e-05, - "loss": 2.2736, + "epoch": 0.28984240304995024, + "grad_norm": 0.95703125, + "learning_rate": 0.0001788613783919812, + "loss": 2.4065, "step": 746 }, { - "epoch": 0.23923138510808648, - "grad_norm": 0.384765625, - "learning_rate": 4.711236512379612e-05, - "loss": 2.1532, + "epoch": 0.2902309317403657, + "grad_norm": 0.74609375, + "learning_rate": 0.00017877786172824952, + "loss": 2.399, "step": 747 }, { - "epoch": 0.23955164131305043, - "grad_norm": 0.70703125, - "learning_rate": 4.709930651816258e-05, - "loss": 2.0565, + "epoch": 0.2906194604307812, + "grad_norm": 0.82421875, + "learning_rate": 0.0001786941999864121, + "loss": 2.4409, "step": 748 }, { - "epoch": 0.2398718975180144, - "grad_norm": 0.671875, - "learning_rate": 4.708622027018504e-05, - "loss": 2.5243, + "epoch": 0.2910079891211967, + "grad_norm": 0.87890625, + "learning_rate": 0.0001786103933205412, + "loss": 2.3703, "step": 749 }, { - "epoch": 0.2401921537229784, - "grad_norm": 0.59765625, - "learning_rate": 4.70731063962321e-05, - "loss": 2.1374, + "epoch": 0.29139651781161213, + "grad_norm": 0.97265625, + "learning_rate": 0.000178526441884976, + "loss": 2.389, "step": 750 }, { - "epoch": 0.24051240992794234, - "grad_norm": 0.52734375, - "learning_rate": 4.705996491270689e-05, - "loss": 2.0636, + "epoch": 0.29178504650202763, + "grad_norm": 0.8984375, + "learning_rate": 0.00017844234583432232, + "loss": 2.4237, "step": 751 }, { - "epoch": 0.24083266613290633, - "grad_norm": 0.8515625, - "learning_rate": 4.704679583604711e-05, - "loss": 2.4744, + "epoch": 0.29217357519244314, + "grad_norm": 0.83984375, + "learning_rate": 0.00017835810532345231, + "loss": 2.4474, "step": 752 }, { - "epoch": 0.2411529223378703, - "grad_norm": 0.6171875, - "learning_rate": 4.703359918272495e-05, - "loss": 2.2096, + "epoch": 0.2925621038828586, + "grad_norm": 0.8671875, + "learning_rate": 0.00017827372050750414, + "loss": 2.3757, "step": 753 }, { - "epoch": 0.24147317854283426, - "grad_norm": 0.58203125, - "learning_rate": 4.7020374969247106e-05, - "loss": 2.576, + "epoch": 0.2929506325732741, + "grad_norm": 0.83203125, + "learning_rate": 0.00017818919154188179, + "loss": 2.4017, "step": 754 }, { - "epoch": 0.24179343474779824, - "grad_norm": 0.6953125, - "learning_rate": 4.700712321215474e-05, - "loss": 2.3694, + "epoch": 0.2933391612636896, + "grad_norm": 0.87109375, + "learning_rate": 0.00017810451858225459, + "loss": 2.4833, "step": 755 }, { - "epoch": 0.24211369095276222, - "grad_norm": 0.451171875, - "learning_rate": 4.699384392802346e-05, - "loss": 2.4635, + "epoch": 0.29372768995410503, + "grad_norm": 0.83203125, + "learning_rate": 0.0001780197017845571, + "loss": 2.4328, "step": 756 }, { - "epoch": 0.24243394715772618, - "grad_norm": 0.5625, - "learning_rate": 4.698053713346334e-05, - "loss": 2.7555, + "epoch": 0.29411621864452053, + "grad_norm": 0.78515625, + "learning_rate": 0.0001779347413049889, + "loss": 2.3754, "step": 757 }, { - "epoch": 0.24275420336269016, - "grad_norm": 0.6640625, - "learning_rate": 4.696720284511881e-05, - "loss": 2.4657, + "epoch": 0.29450474733493603, + "grad_norm": 0.77734375, + "learning_rate": 0.00017784963730001395, + "loss": 2.4475, "step": 758 }, { - "epoch": 0.2430744595676541, - "grad_norm": 1.2734375, - "learning_rate": 4.6953841079668724e-05, - "loss": 1.6963, + "epoch": 0.2948932760253515, + "grad_norm": 0.7890625, + "learning_rate": 0.00017776438992636072, + "loss": 2.4114, "step": 759 }, { - "epoch": 0.2433947157726181, - "grad_norm": 0.439453125, - "learning_rate": 4.6940451853826316e-05, - "loss": 2.1481, + "epoch": 0.295281804715767, + "grad_norm": 0.78515625, + "learning_rate": 0.00017767899934102154, + "loss": 2.4167, "step": 760 }, { - "epoch": 0.24371497197758207, - "grad_norm": 1.0546875, - "learning_rate": 4.692703518433913e-05, - "loss": 2.1694, + "epoch": 0.2956703334061825, + "grad_norm": 0.8046875, + "learning_rate": 0.0001775934657012527, + "loss": 2.4213, "step": 761 }, { - "epoch": 0.24403522818254603, - "grad_norm": 0.47265625, - "learning_rate": 4.6913591087989084e-05, - "loss": 2.0632, + "epoch": 0.29605886209659793, + "grad_norm": 0.87109375, + "learning_rate": 0.00017750778916457373, + "loss": 2.3988, "step": 762 }, { - "epoch": 0.24435548438751, - "grad_norm": 0.453125, - "learning_rate": 4.690011958159235e-05, - "loss": 2.0139, + "epoch": 0.29644739078701343, + "grad_norm": 0.77734375, + "learning_rate": 0.00017742196988876748, + "loss": 2.416, "step": 763 }, { - "epoch": 0.244675740592474, - "grad_norm": 0.3671875, - "learning_rate": 4.6886620681999434e-05, - "loss": 2.0248, + "epoch": 0.29683591947742893, + "grad_norm": 0.859375, + "learning_rate": 0.00017733600803187952, + "loss": 2.4384, "step": 764 }, { - "epoch": 0.24499599679743794, - "grad_norm": 0.4921875, - "learning_rate": 4.6873094406095066e-05, - "loss": 2.3785, + "epoch": 0.2972244481678444, + "grad_norm": 0.83984375, + "learning_rate": 0.0001772499037522181, + "loss": 2.3419, "step": 765 }, { - "epoch": 0.24531625300240192, - "grad_norm": 0.515625, - "learning_rate": 4.685954077079825e-05, - "loss": 2.6487, + "epoch": 0.2976129768582599, + "grad_norm": 0.7421875, + "learning_rate": 0.0001771636572083537, + "loss": 2.4816, "step": 766 }, { - "epoch": 0.2456365092073659, - "grad_norm": 0.953125, - "learning_rate": 4.68459597930622e-05, - "loss": 2.0873, + "epoch": 0.2980015055486754, + "grad_norm": 0.87890625, + "learning_rate": 0.00017707726855911888, + "loss": 2.4267, "step": 767 }, { - "epoch": 0.24595676541232986, - "grad_norm": 0.53515625, - "learning_rate": 4.683235148987432e-05, - "loss": 2.0824, + "epoch": 0.29839003423909083, + "grad_norm": 0.80859375, + "learning_rate": 0.0001769907379636078, + "loss": 2.3589, "step": 768 }, { - "epoch": 0.24627702161729384, - "grad_norm": 0.41796875, - "learning_rate": 4.68187158782562e-05, - "loss": 2.6039, + "epoch": 0.29877856292950633, + "grad_norm": 0.8359375, + "learning_rate": 0.00017690406558117615, + "loss": 2.4548, "step": 769 }, { - "epoch": 0.2465972778222578, - "grad_norm": 0.71484375, - "learning_rate": 4.680505297526361e-05, - "loss": 2.1172, + "epoch": 0.29916709161992183, + "grad_norm": 0.83984375, + "learning_rate": 0.0001768172515714406, + "loss": 2.4251, "step": 770 }, { - "epoch": 0.24691753402722177, - "grad_norm": 0.5546875, - "learning_rate": 4.679136279798642e-05, - "loss": 2.3141, + "epoch": 0.2995556203103373, + "grad_norm": 0.828125, + "learning_rate": 0.0001767302960942787, + "loss": 2.3883, "step": 771 }, { - "epoch": 0.24723779023218576, - "grad_norm": 0.474609375, - "learning_rate": 4.677764536354864e-05, - "loss": 2.0481, + "epoch": 0.2999441490007528, + "grad_norm": 0.87109375, + "learning_rate": 0.00017664319930982863, + "loss": 2.4494, "step": 772 }, { - "epoch": 0.2475580464371497, - "grad_norm": 0.376953125, - "learning_rate": 4.6763900689108365e-05, - "loss": 1.7297, + "epoch": 0.3003326776911683, + "grad_norm": 0.8046875, + "learning_rate": 0.00017655596137848857, + "loss": 2.4177, "step": 773 }, { - "epoch": 0.2478783026421137, - "grad_norm": 1.484375, - "learning_rate": 4.675012879185777e-05, - "loss": 2.4405, + "epoch": 0.3007212063815837, + "grad_norm": 0.8515625, + "learning_rate": 0.00017646858246091692, + "loss": 2.393, "step": 774 }, { - "epoch": 0.24819855884707767, - "grad_norm": 0.53125, - "learning_rate": 4.673632968902306e-05, - "loss": 2.1231, + "epoch": 0.30110973507199923, + "grad_norm": 0.90625, + "learning_rate": 0.00017638106271803143, + "loss": 2.5158, "step": 775 }, { - "epoch": 0.24851881505204163, - "grad_norm": 0.53125, - "learning_rate": 4.67225033978645e-05, - "loss": 1.7979, + "epoch": 0.30149826376241473, + "grad_norm": 0.8359375, + "learning_rate": 0.00017629340231100943, + "loss": 2.4587, "step": 776 }, { - "epoch": 0.2488390712570056, - "grad_norm": 0.5, - "learning_rate": 4.670864993567635e-05, - "loss": 1.7725, + "epoch": 0.3018867924528302, + "grad_norm": 1.0859375, + "learning_rate": 0.00017620560140128722, + "loss": 2.4661, "step": 777 }, { - "epoch": 0.2491593274619696, - "grad_norm": 0.396484375, - "learning_rate": 4.6694769319786834e-05, - "loss": 2.0591, + "epoch": 0.3022753211432457, + "grad_norm": 0.8671875, + "learning_rate": 0.0001761176601505598, + "loss": 2.4737, "step": 778 }, { - "epoch": 0.24947958366693354, - "grad_norm": 0.47265625, - "learning_rate": 4.668086156755818e-05, - "loss": 2.1371, + "epoch": 0.3026638498336612, + "grad_norm": 0.9296875, + "learning_rate": 0.0001760295787207807, + "loss": 2.4462, "step": 779 }, { - "epoch": 0.24979983987189752, - "grad_norm": 0.37109375, - "learning_rate": 4.666692669638653e-05, - "loss": 2.1823, + "epoch": 0.3030523785240766, + "grad_norm": 0.765625, + "learning_rate": 0.00017594135727416155, + "loss": 2.4242, "step": 780 }, { - "epoch": 0.2501200960768615, - "grad_norm": 0.734375, - "learning_rate": 4.665296472370195e-05, - "loss": 1.8962, + "epoch": 0.30344090721449213, + "grad_norm": 1.0234375, + "learning_rate": 0.00017585299597317185, + "loss": 2.4995, "step": 781 }, { - "epoch": 0.2504403522818255, - "grad_norm": 0.40625, - "learning_rate": 4.663897566696843e-05, - "loss": 2.6859, + "epoch": 0.30382943590490763, + "grad_norm": 0.8515625, + "learning_rate": 0.00017576449498053867, + "loss": 2.4151, "step": 782 }, { - "epoch": 0.25076060848678944, - "grad_norm": 0.671875, - "learning_rate": 4.66249595436838e-05, - "loss": 2.4368, + "epoch": 0.3042179645953231, + "grad_norm": 0.8515625, + "learning_rate": 0.00017567585445924633, + "loss": 2.3866, "step": 783 }, { - "epoch": 0.2510808646917534, - "grad_norm": 0.419921875, - "learning_rate": 4.661091637137977e-05, - "loss": 1.895, + "epoch": 0.3046064932857386, + "grad_norm": 0.90625, + "learning_rate": 0.00017558707457253613, + "loss": 2.4381, "step": 784 }, { - "epoch": 0.2514011208967174, - "grad_norm": 0.369140625, - "learning_rate": 4.659684616762188e-05, - "loss": 2.1725, + "epoch": 0.3049950219761541, + "grad_norm": 0.83203125, + "learning_rate": 0.000175498155483906, + "loss": 2.4386, "step": 785 }, { - "epoch": 0.25172137710168135, - "grad_norm": 0.54296875, - "learning_rate": 4.6582748950009475e-05, - "loss": 2.7823, + "epoch": 0.3053835506665695, + "grad_norm": 0.8203125, + "learning_rate": 0.0001754090973571102, + "loss": 2.4855, "step": 786 }, { - "epoch": 0.2520416333066453, - "grad_norm": 0.92578125, - "learning_rate": 4.65686247361757e-05, - "loss": 2.3225, + "epoch": 0.305772079356985, + "grad_norm": 0.80859375, + "learning_rate": 0.00017531990035615915, + "loss": 2.3814, "step": 787 }, { - "epoch": 0.25236188951160926, - "grad_norm": 0.48828125, - "learning_rate": 4.655447354378745e-05, - "loss": 2.1723, + "epoch": 0.3061606080474005, + "grad_norm": 0.8046875, + "learning_rate": 0.00017523056464531885, + "loss": 2.4152, "step": 788 }, { - "epoch": 0.25268214571657327, - "grad_norm": 0.412109375, - "learning_rate": 4.654029539054539e-05, - "loss": 1.9019, + "epoch": 0.306549136737816, + "grad_norm": 0.87890625, + "learning_rate": 0.0001751410903891109, + "loss": 2.4139, "step": 789 }, { - "epoch": 0.2530024019215372, - "grad_norm": 0.451171875, - "learning_rate": 4.652609029418389e-05, - "loss": 2.178, + "epoch": 0.3069376654282315, + "grad_norm": 0.83203125, + "learning_rate": 0.000175051477752312, + "loss": 2.4099, "step": 790 }, { - "epoch": 0.2533226581265012, - "grad_norm": 1.03125, - "learning_rate": 4.651185827247103e-05, - "loss": 1.4949, + "epoch": 0.3073261941186469, + "grad_norm": 0.84765625, + "learning_rate": 0.00017496172689995368, + "loss": 2.4541, "step": 791 }, { - "epoch": 0.2536429143314652, - "grad_norm": 0.58984375, - "learning_rate": 4.649759934320855e-05, - "loss": 2.6352, + "epoch": 0.3077147228090624, + "grad_norm": 0.84375, + "learning_rate": 0.00017487183799732203, + "loss": 2.5013, "step": 792 }, { - "epoch": 0.25396317053642914, - "grad_norm": 0.431640625, - "learning_rate": 4.6483313524231874e-05, - "loss": 2.1695, + "epoch": 0.3081032514994779, + "grad_norm": 0.796875, + "learning_rate": 0.0001747818112099573, + "loss": 2.413, "step": 793 }, { - "epoch": 0.2542834267413931, - "grad_norm": 0.50390625, - "learning_rate": 4.646900083341005e-05, - "loss": 2.4847, + "epoch": 0.30849178018989337, + "grad_norm": 0.80859375, + "learning_rate": 0.0001746916467036538, + "loss": 2.4217, "step": 794 }, { - "epoch": 0.2546036829463571, - "grad_norm": 0.5703125, - "learning_rate": 4.645466128864573e-05, - "loss": 2.0651, + "epoch": 0.3088803088803089, + "grad_norm": 0.87109375, + "learning_rate": 0.00017460134464445935, + "loss": 2.4751, "step": 795 }, { - "epoch": 0.25492393915132106, - "grad_norm": 0.4765625, - "learning_rate": 4.644029490787517e-05, - "loss": 2.7276, + "epoch": 0.3092688375707244, + "grad_norm": 0.875, + "learning_rate": 0.00017451090519867517, + "loss": 2.4224, "step": 796 }, { - "epoch": 0.255244195356285, - "grad_norm": 0.5625, - "learning_rate": 4.642590170906816e-05, - "loss": 2.5108, + "epoch": 0.3096573662611398, + "grad_norm": 0.81640625, + "learning_rate": 0.00017442032853285543, + "loss": 2.3641, "step": 797 }, { - "epoch": 0.255564451561249, - "grad_norm": 0.44921875, - "learning_rate": 4.6411481710228096e-05, - "loss": 2.3693, + "epoch": 0.3100458949515553, + "grad_norm": 0.94921875, + "learning_rate": 0.00017432961481380707, + "loss": 2.5269, "step": 798 }, { - "epoch": 0.25588470776621297, - "grad_norm": 0.4765625, - "learning_rate": 4.6397034929391834e-05, - "loss": 2.333, + "epoch": 0.3104344236419708, + "grad_norm": 0.76953125, + "learning_rate": 0.00017423876420858932, + "loss": 2.3866, "step": 799 }, { - "epoch": 0.2562049639711769, - "grad_norm": 0.5078125, - "learning_rate": 4.638256138462975e-05, - "loss": 2.4582, + "epoch": 0.31082295233238627, + "grad_norm": 0.7890625, + "learning_rate": 0.00017414777688451368, + "loss": 2.4445, "step": 800 }, { - "epoch": 0.25652522017614093, - "grad_norm": 0.365234375, - "learning_rate": 4.636806109404573e-05, - "loss": 2.5472, + "epoch": 0.3112114810228018, + "grad_norm": 0.91015625, + "learning_rate": 0.0001740566530091432, + "loss": 2.4237, "step": 801 }, { - "epoch": 0.2568454763811049, - "grad_norm": 0.5078125, - "learning_rate": 4.635353407577706e-05, - "loss": 2.4438, + "epoch": 0.3116000097132173, + "grad_norm": 0.87109375, + "learning_rate": 0.00017396539275029262, + "loss": 2.4056, "step": 802 }, { - "epoch": 0.25716573258606884, - "grad_norm": 0.44140625, - "learning_rate": 4.633898034799449e-05, - "loss": 2.3358, + "epoch": 0.3119885384036327, + "grad_norm": 0.91796875, + "learning_rate": 0.00017387399627602772, + "loss": 2.4517, "step": 803 }, { - "epoch": 0.25748598879103285, - "grad_norm": 0.921875, - "learning_rate": 4.632439992890218e-05, - "loss": 2.4705, + "epoch": 0.3123770670940482, + "grad_norm": 0.9375, + "learning_rate": 0.00017378246375466513, + "loss": 2.4313, "step": 804 }, { - "epoch": 0.2578062449959968, - "grad_norm": 0.3984375, - "learning_rate": 4.630979283673766e-05, - "loss": 2.163, + "epoch": 0.3127655957844637, + "grad_norm": 0.8203125, + "learning_rate": 0.00017369079535477217, + "loss": 2.4828, "step": 805 }, { - "epoch": 0.25812650120096076, - "grad_norm": 0.51953125, - "learning_rate": 4.629515908977184e-05, - "loss": 1.7262, + "epoch": 0.31315412447487917, + "grad_norm": 0.89453125, + "learning_rate": 0.00017359899124516623, + "loss": 2.4158, "step": 806 }, { - "epoch": 0.25844675740592477, - "grad_norm": 0.4296875, - "learning_rate": 4.628049870630896e-05, - "loss": 2.2349, + "epoch": 0.31354265316529467, + "grad_norm": 0.921875, + "learning_rate": 0.00017350705159491464, + "loss": 2.3548, "step": 807 }, { - "epoch": 0.2587670136108887, - "grad_norm": 0.431640625, - "learning_rate": 4.626581170468658e-05, - "loss": 2.3561, + "epoch": 0.3139311818557102, + "grad_norm": 0.83984375, + "learning_rate": 0.00017341497657333448, + "loss": 2.427, "step": 808 }, { - "epoch": 0.25908726981585267, - "grad_norm": 0.458984375, - "learning_rate": 4.625109810327556e-05, - "loss": 2.5208, + "epoch": 0.3143197105461256, + "grad_norm": 0.76171875, + "learning_rate": 0.000173322766349992, + "loss": 2.4104, "step": 809 }, { - "epoch": 0.2594075260208166, - "grad_norm": 1.953125, - "learning_rate": 4.6236357920480013e-05, - "loss": 2.0921, + "epoch": 0.3147082392365411, + "grad_norm": 0.859375, + "learning_rate": 0.00017323042109470245, + "loss": 2.4454, "step": 810 }, { - "epoch": 0.25972778222578063, - "grad_norm": 0.828125, - "learning_rate": 4.6221591174737314e-05, - "loss": 2.2984, + "epoch": 0.3150967679269566, + "grad_norm": 0.91796875, + "learning_rate": 0.0001731379409775298, + "loss": 2.4989, "step": 811 }, { - "epoch": 0.2600480384307446, - "grad_norm": 0.3359375, - "learning_rate": 4.620679788451808e-05, - "loss": 1.6466, + "epoch": 0.31548529661737207, + "grad_norm": 0.87890625, + "learning_rate": 0.0001730453261687864, + "loss": 2.4115, "step": 812 }, { - "epoch": 0.26036829463570854, - "grad_norm": 0.427734375, - "learning_rate": 4.61919780683261e-05, - "loss": 2.3331, + "epoch": 0.31587382530778757, + "grad_norm": 0.84375, + "learning_rate": 0.00017295257683903257, + "loss": 2.3898, "step": 813 }, { - "epoch": 0.26068855084067255, - "grad_norm": 0.404296875, - "learning_rate": 4.6177131744698364e-05, - "loss": 2.313, + "epoch": 0.31626235399820307, + "grad_norm": 0.96875, + "learning_rate": 0.00017285969315907648, + "loss": 2.3612, "step": 814 }, { - "epoch": 0.2610088070456365, - "grad_norm": 0.5703125, - "learning_rate": 4.616225893220502e-05, - "loss": 2.4781, + "epoch": 0.3166508826886185, + "grad_norm": 0.86328125, + "learning_rate": 0.00017276667529997355, + "loss": 2.4465, "step": 815 }, { - "epoch": 0.26132906325060046, - "grad_norm": 0.447265625, - "learning_rate": 4.6147359649449305e-05, - "loss": 2.2835, + "epoch": 0.317039411379034, + "grad_norm": 0.80859375, + "learning_rate": 0.0001726735234330265, + "loss": 2.4731, "step": 816 }, { - "epoch": 0.26164931945556447, - "grad_norm": 0.443359375, - "learning_rate": 4.613243391506764e-05, - "loss": 2.2348, + "epoch": 0.3174279400694495, + "grad_norm": 0.82421875, + "learning_rate": 0.0001725802377297847, + "loss": 2.4403, "step": 817 }, { - "epoch": 0.2619695756605284, - "grad_norm": 0.345703125, - "learning_rate": 4.6117481747729476e-05, - "loss": 2.4842, + "epoch": 0.31781646875986497, + "grad_norm": 0.89453125, + "learning_rate": 0.0001724868183620441, + "loss": 2.3637, "step": 818 }, { - "epoch": 0.2622898318654924, - "grad_norm": 0.51953125, - "learning_rate": 4.6102503166137336e-05, - "loss": 2.0446, + "epoch": 0.31820499745028047, + "grad_norm": 0.78515625, + "learning_rate": 0.00017239326550184668, + "loss": 2.3778, "step": 819 }, { - "epoch": 0.2626100880704564, - "grad_norm": 0.46875, - "learning_rate": 4.608749818902679e-05, - "loss": 1.8785, + "epoch": 0.31859352614069597, + "grad_norm": 0.8046875, + "learning_rate": 0.00017229957932148035, + "loss": 2.451, "step": 820 }, { - "epoch": 0.26293034427542034, - "grad_norm": 0.76953125, - "learning_rate": 4.607246683516644e-05, - "loss": 2.5239, + "epoch": 0.3189820548311114, + "grad_norm": 0.79296875, + "learning_rate": 0.00017220575999347856, + "loss": 2.3804, "step": 821 }, { - "epoch": 0.2632506004803843, - "grad_norm": 0.49609375, - "learning_rate": 4.605740912335786e-05, - "loss": 2.61, + "epoch": 0.3193705835215269, + "grad_norm": 0.84765625, + "learning_rate": 0.0001721118076906199, + "loss": 2.3504, "step": 822 }, { - "epoch": 0.2635708566853483, - "grad_norm": 0.48046875, - "learning_rate": 4.604232507243559e-05, - "loss": 1.8585, + "epoch": 0.3197591122119424, + "grad_norm": 0.796875, + "learning_rate": 0.0001720177225859279, + "loss": 2.4297, "step": 823 }, { - "epoch": 0.26389111289031225, - "grad_norm": 0.73828125, - "learning_rate": 4.6027214701267126e-05, - "loss": 2.3605, + "epoch": 0.32014764090235787, + "grad_norm": 0.78515625, + "learning_rate": 0.00017192350485267064, + "loss": 2.3792, "step": 824 }, { - "epoch": 0.2642113690952762, - "grad_norm": 0.3828125, - "learning_rate": 4.6012078028752885e-05, - "loss": 2.2571, + "epoch": 0.32053616959277337, + "grad_norm": 0.82421875, + "learning_rate": 0.00017182915466436045, + "loss": 2.3826, "step": 825 }, { - "epoch": 0.2645316253002402, - "grad_norm": 0.40625, - "learning_rate": 4.5996915073826187e-05, - "loss": 1.8091, + "epoch": 0.32092469828318887, + "grad_norm": 0.7890625, + "learning_rate": 0.00017173467219475352, + "loss": 2.3815, "step": 826 }, { - "epoch": 0.26485188150520417, - "grad_norm": 0.45703125, - "learning_rate": 4.598172585545321e-05, - "loss": 2.3873, + "epoch": 0.3213132269736043, + "grad_norm": 0.82421875, + "learning_rate": 0.00017164005761784984, + "loss": 2.4616, "step": 827 }, { - "epoch": 0.2651721377101681, - "grad_norm": 0.5703125, - "learning_rate": 4.5966510392633e-05, - "loss": 2.171, + "epoch": 0.3217017556640198, + "grad_norm": 0.84765625, + "learning_rate": 0.00017154531110789248, + "loss": 2.3989, "step": 828 }, { - "epoch": 0.26549239391513213, - "grad_norm": 2.75, - "learning_rate": 4.595126870439742e-05, - "loss": 2.4189, + "epoch": 0.3220902843544353, + "grad_norm": 0.86328125, + "learning_rate": 0.0001714504328393676, + "loss": 2.4294, "step": 829 }, { - "epoch": 0.2658126501200961, - "grad_norm": 0.482421875, - "learning_rate": 4.593600080981114e-05, - "loss": 2.3643, + "epoch": 0.32247881304485077, + "grad_norm": 0.89453125, + "learning_rate": 0.00017135542298700397, + "loss": 2.4438, "step": 830 }, { - "epoch": 0.26613290632506004, - "grad_norm": 0.52734375, - "learning_rate": 4.592070672797161e-05, - "loss": 2.1314, + "epoch": 0.32286734173526627, + "grad_norm": 0.76953125, + "learning_rate": 0.00017126028172577274, + "loss": 2.4648, "step": 831 }, { - "epoch": 0.26645316253002405, - "grad_norm": 0.6328125, - "learning_rate": 4.590538647800904e-05, - "loss": 2.5095, + "epoch": 0.32325587042568177, + "grad_norm": 0.8984375, + "learning_rate": 0.00017116500923088697, + "loss": 2.5006, "step": 832 }, { - "epoch": 0.266773418734988, - "grad_norm": 0.6015625, - "learning_rate": 4.5890040079086374e-05, - "loss": 2.1051, + "epoch": 0.3236443991160972, + "grad_norm": 0.84765625, + "learning_rate": 0.0001710696056778014, + "loss": 2.4287, "step": 833 }, { - "epoch": 0.26709367493995195, - "grad_norm": 0.54296875, - "learning_rate": 4.587466755039924e-05, - "loss": 2.6451, + "epoch": 0.3240329278065127, + "grad_norm": 0.83984375, + "learning_rate": 0.0001709740712422123, + "loss": 2.3856, "step": 834 }, { - "epoch": 0.2674139311449159, - "grad_norm": 0.486328125, - "learning_rate": 4.585926891117597e-05, - "loss": 2.3557, + "epoch": 0.3244214564969282, + "grad_norm": 0.80078125, + "learning_rate": 0.00017087840610005675, + "loss": 2.4137, "step": 835 }, { - "epoch": 0.2677341873498799, - "grad_norm": 0.435546875, - "learning_rate": 4.584384418067756e-05, - "loss": 2.2055, + "epoch": 0.32480998518734366, + "grad_norm": 0.921875, + "learning_rate": 0.0001707826104275127, + "loss": 2.5085, "step": 836 }, { - "epoch": 0.26805444355484387, - "grad_norm": 0.55859375, - "learning_rate": 4.582839337819763e-05, - "loss": 1.711, + "epoch": 0.32519851387775917, + "grad_norm": 0.85546875, + "learning_rate": 0.0001706866844009984, + "loss": 2.3359, "step": 837 }, { - "epoch": 0.2683746997598078, - "grad_norm": 1.6640625, - "learning_rate": 4.581291652306241e-05, - "loss": 2.7919, + "epoch": 0.32558704256817467, + "grad_norm": 0.703125, + "learning_rate": 0.00017059062819717218, + "loss": 2.3356, "step": 838 }, { - "epoch": 0.26869495596477183, - "grad_norm": 0.51171875, - "learning_rate": 4.579741363463074e-05, - "loss": 2.3713, + "epoch": 0.3259755712585901, + "grad_norm": 0.8671875, + "learning_rate": 0.00017049444199293215, + "loss": 2.4243, "step": 839 }, { - "epoch": 0.2690152121697358, - "grad_norm": 0.625, - "learning_rate": 4.578188473229398e-05, - "loss": 2.4531, + "epoch": 0.3263640999490056, + "grad_norm": 0.91015625, + "learning_rate": 0.00017039812596541574, + "loss": 2.4393, "step": 840 }, { - "epoch": 0.26933546837469974, - "grad_norm": 0.48828125, - "learning_rate": 4.576632983547607e-05, - "loss": 2.2186, + "epoch": 0.3267526286394211, + "grad_norm": 0.82421875, + "learning_rate": 0.00017030168029199958, + "loss": 2.3736, "step": 841 }, { - "epoch": 0.26965572457966375, - "grad_norm": 0.51953125, - "learning_rate": 4.575074896363345e-05, - "loss": 2.3781, + "epoch": 0.32714115732983656, + "grad_norm": 0.875, + "learning_rate": 0.00017020510515029894, + "loss": 2.4372, "step": 842 }, { - "epoch": 0.2699759807846277, - "grad_norm": 0.482421875, - "learning_rate": 4.573514213625505e-05, - "loss": 2.1847, + "epoch": 0.32752968602025206, + "grad_norm": 0.88671875, + "learning_rate": 0.00017010840071816764, + "loss": 2.4257, "step": 843 }, { - "epoch": 0.27029623698959165, - "grad_norm": 0.421875, - "learning_rate": 4.571950937286225e-05, - "loss": 2.1921, + "epoch": 0.32791821471066757, + "grad_norm": 0.85546875, + "learning_rate": 0.0001700115671736975, + "loss": 2.3996, "step": 844 }, { - "epoch": 0.27061649319455566, - "grad_norm": 0.3984375, - "learning_rate": 4.570385069300891e-05, - "loss": 2.6008, + "epoch": 0.328306743401083, + "grad_norm": 0.85546875, + "learning_rate": 0.00016991460469521817, + "loss": 2.3186, "step": 845 }, { - "epoch": 0.2709367493995196, - "grad_norm": 0.671875, - "learning_rate": 4.568816611628126e-05, - "loss": 2.5111, + "epoch": 0.3286952720914985, + "grad_norm": 1.15625, + "learning_rate": 0.00016981751346129668, + "loss": 2.4109, "step": 846 }, { - "epoch": 0.27125700560448357, - "grad_norm": 0.412109375, - "learning_rate": 4.5672455662297966e-05, - "loss": 2.4546, + "epoch": 0.329083800781914, + "grad_norm": 0.76953125, + "learning_rate": 0.0001697202936507373, + "loss": 2.4756, "step": 847 }, { - "epoch": 0.2715772618094476, - "grad_norm": 0.490234375, - "learning_rate": 4.565671935071002e-05, - "loss": 2.2381, + "epoch": 0.32947232947232946, + "grad_norm": 0.9375, + "learning_rate": 0.00016962294544258096, + "loss": 2.4485, "step": 848 }, { - "epoch": 0.27189751801441153, - "grad_norm": 0.66015625, - "learning_rate": 4.56409572012008e-05, - "loss": 2.5955, + "epoch": 0.32986085816274496, + "grad_norm": 0.91796875, + "learning_rate": 0.00016952546901610513, + "loss": 2.3982, "step": 849 }, { - "epoch": 0.2722177742193755, - "grad_norm": 0.51171875, - "learning_rate": 4.562516923348597e-05, - "loss": 2.2343, + "epoch": 0.33024938685316046, + "grad_norm": 0.76953125, + "learning_rate": 0.0001694278645508234, + "loss": 2.4209, "step": 850 }, { - "epoch": 0.2725380304243395, - "grad_norm": 0.45703125, - "learning_rate": 4.56093554673135e-05, - "loss": 2.4699, + "epoch": 0.3306379155435759, + "grad_norm": 0.82421875, + "learning_rate": 0.00016933013222648508, + "loss": 2.4256, "step": 851 }, { - "epoch": 0.27285828662930345, - "grad_norm": 0.5703125, - "learning_rate": 4.559351592246363e-05, - "loss": 2.4249, + "epoch": 0.3310264442339914, + "grad_norm": 0.8359375, + "learning_rate": 0.00016923227222307506, + "loss": 2.4372, "step": 852 }, { - "epoch": 0.2731785428342674, - "grad_norm": 0.9765625, - "learning_rate": 4.5577650618748846e-05, - "loss": 1.7957, + "epoch": 0.3314149729244069, + "grad_norm": 0.8046875, + "learning_rate": 0.00016913428472081326, + "loss": 2.366, "step": 853 }, { - "epoch": 0.2734987990392314, - "grad_norm": 0.61328125, - "learning_rate": 4.556175957601385e-05, - "loss": 2.3653, + "epoch": 0.33180350161482236, + "grad_norm": 0.8125, + "learning_rate": 0.00016903616990015453, + "loss": 2.3984, "step": 854 }, { - "epoch": 0.27381905524419536, - "grad_norm": 0.4765625, - "learning_rate": 4.5545842814135545e-05, - "loss": 2.7702, + "epoch": 0.33219203030523786, + "grad_norm": 0.92578125, + "learning_rate": 0.00016893792794178805, + "loss": 2.3797, "step": 855 }, { - "epoch": 0.2741393114491593, - "grad_norm": 0.453125, - "learning_rate": 4.552990035302299e-05, - "loss": 2.3475, + "epoch": 0.3325805589956533, + "grad_norm": 0.82421875, + "learning_rate": 0.0001688395590266372, + "loss": 2.4639, "step": 856 }, { - "epoch": 0.27445956765412327, - "grad_norm": 0.61328125, - "learning_rate": 4.55139322126174e-05, - "loss": 2.1684, + "epoch": 0.3329690876860688, + "grad_norm": 0.828125, + "learning_rate": 0.0001687410633358592, + "loss": 2.378, "step": 857 }, { - "epoch": 0.2747798238590873, - "grad_norm": 0.396484375, - "learning_rate": 4.549793841289211e-05, - "loss": 1.8475, + "epoch": 0.3333576163764843, + "grad_norm": 0.87890625, + "learning_rate": 0.00016864244105084473, + "loss": 2.4141, "step": 858 }, { - "epoch": 0.27510008006405123, - "grad_norm": 1.1953125, - "learning_rate": 4.548191897385257e-05, - "loss": 2.3642, + "epoch": 0.33374614506689976, + "grad_norm": 0.80078125, + "learning_rate": 0.00016854369235321754, + "loss": 2.4147, "step": 859 }, { - "epoch": 0.2754203362690152, - "grad_norm": 0.458984375, - "learning_rate": 4.546587391553624e-05, - "loss": 1.9096, + "epoch": 0.33413467375731526, + "grad_norm": 0.78125, + "learning_rate": 0.00016844481742483424, + "loss": 2.3797, "step": 860 }, { - "epoch": 0.2757405924739792, - "grad_norm": 0.54296875, - "learning_rate": 4.54498032580127e-05, - "loss": 2.4123, + "epoch": 0.33452320244773076, + "grad_norm": 0.8359375, + "learning_rate": 0.0001683458164477839, + "loss": 2.3711, "step": 861 }, { - "epoch": 0.27606084867894315, - "grad_norm": 0.5703125, - "learning_rate": 4.54337070213835e-05, - "loss": 2.431, + "epoch": 0.3349117311381462, + "grad_norm": 0.85546875, + "learning_rate": 0.00016824668960438771, + "loss": 2.4823, "step": 862 }, { - "epoch": 0.2763811048839071, - "grad_norm": 0.49609375, - "learning_rate": 4.54175852257822e-05, - "loss": 2.5574, + "epoch": 0.3353002598285617, + "grad_norm": 0.875, + "learning_rate": 0.00016814743707719868, + "loss": 2.464, "step": 863 }, { - "epoch": 0.2767013610888711, - "grad_norm": 0.46875, - "learning_rate": 4.5401437891374326e-05, - "loss": 2.4494, + "epoch": 0.3356887885189772, + "grad_norm": 0.88671875, + "learning_rate": 0.00016804805904900127, + "loss": 2.4662, "step": 864 }, { - "epoch": 0.27702161729383507, - "grad_norm": 0.416015625, - "learning_rate": 4.5385265038357363e-05, - "loss": 2.2868, + "epoch": 0.33607731720939266, + "grad_norm": 1.0078125, + "learning_rate": 0.00016794855570281106, + "loss": 2.3954, "step": 865 }, { - "epoch": 0.277341873498799, - "grad_norm": 1.3359375, - "learning_rate": 4.5369066686960694e-05, - "loss": 2.2944, + "epoch": 0.33646584589980816, + "grad_norm": 0.9140625, + "learning_rate": 0.00016784892722187438, + "loss": 2.4792, "step": 866 }, { - "epoch": 0.277662129703763, - "grad_norm": 0.7734375, - "learning_rate": 4.535284285744561e-05, - "loss": 1.8372, + "epoch": 0.33685437459022366, + "grad_norm": 0.87890625, + "learning_rate": 0.0001677491737896681, + "loss": 2.3987, "step": 867 }, { - "epoch": 0.277982385908727, - "grad_norm": 0.5, - "learning_rate": 4.5336593570105266e-05, - "loss": 2.3709, + "epoch": 0.3372429032806391, + "grad_norm": 0.78125, + "learning_rate": 0.00016764929558989907, + "loss": 2.3545, "step": 868 }, { - "epoch": 0.27830264211369093, - "grad_norm": 0.443359375, - "learning_rate": 4.5320318845264664e-05, - "loss": 2.1215, + "epoch": 0.3376314319710546, + "grad_norm": 0.91015625, + "learning_rate": 0.0001675492928065041, + "loss": 2.3862, "step": 869 }, { - "epoch": 0.27862289831865494, - "grad_norm": 0.61328125, - "learning_rate": 4.530401870328063e-05, - "loss": 2.8229, + "epoch": 0.3380199606614701, + "grad_norm": 0.84375, + "learning_rate": 0.00016744916562364928, + "loss": 2.4111, "step": 870 }, { - "epoch": 0.2789431545236189, - "grad_norm": 0.9765625, - "learning_rate": 4.528769316454176e-05, - "loss": 2.272, + "epoch": 0.33840848935188556, + "grad_norm": 0.8828125, + "learning_rate": 0.0001673489142257298, + "loss": 2.4104, "step": 871 }, { - "epoch": 0.27926341072858285, - "grad_norm": 0.546875, - "learning_rate": 4.5271342249468424e-05, - "loss": 2.1168, + "epoch": 0.33879701804230106, + "grad_norm": 0.8046875, + "learning_rate": 0.0001672485387973697, + "loss": 2.3908, "step": 872 }, { - "epoch": 0.27958366693354686, - "grad_norm": 0.33203125, - "learning_rate": 4.525496597851275e-05, - "loss": 1.5178, + "epoch": 0.33918554673271656, + "grad_norm": 0.8203125, + "learning_rate": 0.00016714803952342134, + "loss": 2.3626, "step": 873 }, { - "epoch": 0.2799039231385108, - "grad_norm": 0.44140625, - "learning_rate": 4.523856437215858e-05, - "loss": 1.6794, + "epoch": 0.339574075423132, + "grad_norm": 0.84765625, + "learning_rate": 0.00016704741658896526, + "loss": 2.4738, "step": 874 }, { - "epoch": 0.28022417934347477, - "grad_norm": 0.375, - "learning_rate": 4.522213745092143e-05, - "loss": 2.3376, + "epoch": 0.3399626041135475, + "grad_norm": 0.80859375, + "learning_rate": 0.0001669466701793096, + "loss": 2.515, "step": 875 }, { - "epoch": 0.2805444355484388, - "grad_norm": 0.62890625, - "learning_rate": 4.520568523534848e-05, - "loss": 2.4695, + "epoch": 0.340351132803963, + "grad_norm": 0.78125, + "learning_rate": 0.00016684580047999, + "loss": 2.4536, "step": 876 }, { - "epoch": 0.28086469175340273, - "grad_norm": 0.4609375, - "learning_rate": 4.5189207746018566e-05, - "loss": 1.6355, + "epoch": 0.34073966149437845, + "grad_norm": 0.765625, + "learning_rate": 0.0001667448076767691, + "loss": 2.4085, "step": 877 }, { - "epoch": 0.2811849479583667, - "grad_norm": 0.49609375, - "learning_rate": 4.5172705003542125e-05, - "loss": 2.0943, + "epoch": 0.34112819018479396, + "grad_norm": 0.83203125, + "learning_rate": 0.00016664369195563633, + "loss": 2.4197, "step": 878 }, { - "epoch": 0.2815052041633307, - "grad_norm": 0.62109375, - "learning_rate": 4.515617702856119e-05, - "loss": 1.9592, + "epoch": 0.34151671887520946, + "grad_norm": 0.7734375, + "learning_rate": 0.00016654245350280737, + "loss": 2.3588, "step": 879 }, { - "epoch": 0.28182546036829464, - "grad_norm": 0.484375, - "learning_rate": 4.5139623841749346e-05, - "loss": 2.0776, + "epoch": 0.3419052475656249, + "grad_norm": 0.86328125, + "learning_rate": 0.00016644109250472396, + "loss": 2.4536, "step": 880 }, { - "epoch": 0.2821457165732586, - "grad_norm": 0.5859375, - "learning_rate": 4.5123045463811706e-05, - "loss": 2.2985, + "epoch": 0.3422937762560404, + "grad_norm": 0.796875, + "learning_rate": 0.00016633960914805358, + "loss": 2.4565, "step": 881 }, { - "epoch": 0.28246597277822255, - "grad_norm": 0.431640625, - "learning_rate": 4.510644191548492e-05, - "loss": 1.9178, + "epoch": 0.3426823049464559, + "grad_norm": 0.7421875, + "learning_rate": 0.00016623800361968894, + "loss": 2.4165, "step": 882 }, { - "epoch": 0.28278622898318656, - "grad_norm": 0.3984375, - "learning_rate": 4.50898132175371e-05, - "loss": 2.3869, + "epoch": 0.34307083363687135, + "grad_norm": 0.74609375, + "learning_rate": 0.0001661362761067479, + "loss": 2.3367, "step": 883 }, { - "epoch": 0.2831064851881505, - "grad_norm": 0.625, - "learning_rate": 4.507315939076782e-05, - "loss": 1.5066, + "epoch": 0.34345936232728685, + "grad_norm": 0.7890625, + "learning_rate": 0.0001660344267965728, + "loss": 2.3624, "step": 884 }, { - "epoch": 0.28342674139311447, - "grad_norm": 0.435546875, - "learning_rate": 4.5056480456008085e-05, - "loss": 2.5622, + "epoch": 0.34384789101770236, + "grad_norm": 0.75, + "learning_rate": 0.0001659324558767304, + "loss": 2.4241, "step": 885 }, { - "epoch": 0.2837469975980785, - "grad_norm": 1.0546875, - "learning_rate": 4.5039776434120306e-05, - "loss": 2.3004, + "epoch": 0.3442364197081178, + "grad_norm": 0.80859375, + "learning_rate": 0.00016583036353501134, + "loss": 2.4773, "step": 886 }, { - "epoch": 0.28406725380304243, - "grad_norm": 0.5078125, - "learning_rate": 4.502304734599828e-05, - "loss": 2.3096, + "epoch": 0.3446249483985333, + "grad_norm": 0.7578125, + "learning_rate": 0.00016572814995942993, + "loss": 2.4632, "step": 887 }, { - "epoch": 0.2843875100080064, - "grad_norm": 0.6015625, - "learning_rate": 4.5006293212567164e-05, - "loss": 2.1645, + "epoch": 0.3450134770889488, + "grad_norm": 0.7578125, + "learning_rate": 0.00016562581533822375, + "loss": 2.3947, "step": 888 }, { - "epoch": 0.2847077662129704, - "grad_norm": 0.5625, - "learning_rate": 4.4989514054783406e-05, - "loss": 2.2932, + "epoch": 0.34540200577936425, + "grad_norm": 0.796875, + "learning_rate": 0.0001655233598598532, + "loss": 2.4371, "step": 889 }, { - "epoch": 0.28502802241793435, - "grad_norm": 0.462890625, - "learning_rate": 4.49727098936348e-05, - "loss": 2.1796, + "epoch": 0.34579053446977975, + "grad_norm": 0.734375, + "learning_rate": 0.0001654207837130014, + "loss": 2.377, "step": 890 }, { - "epoch": 0.2853482786228983, - "grad_norm": 0.404296875, - "learning_rate": 4.495588075014038e-05, - "loss": 1.7093, + "epoch": 0.34617906316019526, + "grad_norm": 0.765625, + "learning_rate": 0.0001653180870865736, + "loss": 2.3894, "step": 891 }, { - "epoch": 0.2856685348278623, - "grad_norm": 0.4921875, - "learning_rate": 4.4939026645350454e-05, - "loss": 2.5258, + "epoch": 0.3465675918506107, + "grad_norm": 0.83984375, + "learning_rate": 0.00016521527016969689, + "loss": 2.427, "step": 892 }, { - "epoch": 0.28598879103282626, - "grad_norm": 2.9375, - "learning_rate": 4.4922147600346545e-05, - "loss": 1.979, + "epoch": 0.3469561205410262, + "grad_norm": 0.78125, + "learning_rate": 0.00016511233315172, + "loss": 2.4122, "step": 893 }, { - "epoch": 0.2863090472377902, - "grad_norm": 0.361328125, - "learning_rate": 4.4905243636241354e-05, - "loss": 2.0999, + "epoch": 0.3473446492314417, + "grad_norm": 0.78515625, + "learning_rate": 0.00016500927622221275, + "loss": 2.4176, "step": 894 }, { - "epoch": 0.2866293034427542, - "grad_norm": 0.53515625, - "learning_rate": 4.4888314774178776e-05, - "loss": 2.4738, + "epoch": 0.34773317792185715, + "grad_norm": 0.765625, + "learning_rate": 0.00016490609957096588, + "loss": 2.4046, "step": 895 }, { - "epoch": 0.2869495596477182, - "grad_norm": 0.8046875, - "learning_rate": 4.4871361035333836e-05, - "loss": 2.3017, + "epoch": 0.34812170661227265, + "grad_norm": 0.84375, + "learning_rate": 0.0001648028033879905, + "loss": 2.3702, "step": 896 }, { - "epoch": 0.28726981585268213, - "grad_norm": 0.384765625, - "learning_rate": 4.485438244091268e-05, - "loss": 2.2876, + "epoch": 0.34851023530268815, + "grad_norm": 0.82421875, + "learning_rate": 0.00016469938786351786, + "loss": 2.4124, "step": 897 }, { - "epoch": 0.28759007205764614, - "grad_norm": 0.458984375, - "learning_rate": 4.483737901215254e-05, - "loss": 2.1632, + "epoch": 0.3488987639931036, + "grad_norm": 0.7734375, + "learning_rate": 0.00016459585318799914, + "loss": 2.3884, "step": 898 }, { - "epoch": 0.2879103282626101, - "grad_norm": 0.54296875, - "learning_rate": 4.482035077032171e-05, - "loss": 2.4111, + "epoch": 0.3492872926835191, + "grad_norm": 0.765625, + "learning_rate": 0.00016449219955210476, + "loss": 2.38, "step": 899 }, { - "epoch": 0.28823058446757405, - "grad_norm": 0.36328125, - "learning_rate": 4.4803297736719534e-05, - "loss": 2.6698, + "epoch": 0.3496758213739346, + "grad_norm": 0.86328125, + "learning_rate": 0.00016438842714672436, + "loss": 2.4462, "step": 900 }, { - "epoch": 0.28855084067253806, - "grad_norm": 1.5390625, - "learning_rate": 4.478621993267635e-05, - "loss": 2.4385, + "epoch": 0.35006435006435005, + "grad_norm": 0.78125, + "learning_rate": 0.0001642845361629662, + "loss": 2.39, "step": 901 }, { - "epoch": 0.288871096877502, - "grad_norm": 1.1953125, - "learning_rate": 4.476911737955349e-05, - "loss": 1.9053, + "epoch": 0.35045287875476555, + "grad_norm": 0.77734375, + "learning_rate": 0.000164180526792157, + "loss": 2.4532, "step": 902 }, { - "epoch": 0.28919135308246596, - "grad_norm": 0.56640625, - "learning_rate": 4.475199009874323e-05, - "loss": 1.8812, + "epoch": 0.35084140744518105, + "grad_norm": 0.9140625, + "learning_rate": 0.0001640763992258415, + "loss": 2.4183, "step": 903 }, { - "epoch": 0.2895116092874299, - "grad_norm": 0.40234375, - "learning_rate": 4.47348381116688e-05, - "loss": 1.812, + "epoch": 0.3512299361355965, + "grad_norm": 0.828125, + "learning_rate": 0.00016397215365578197, + "loss": 2.4412, "step": 904 }, { - "epoch": 0.2898318654923939, - "grad_norm": 0.50390625, - "learning_rate": 4.47176614397843e-05, - "loss": 2.3822, + "epoch": 0.351618464826012, + "grad_norm": 0.83984375, + "learning_rate": 0.00016386779027395822, + "loss": 2.4002, "step": 905 }, { - "epoch": 0.2901521216973579, - "grad_norm": 0.8828125, - "learning_rate": 4.470046010457473e-05, - "loss": 1.8144, + "epoch": 0.3520069935164275, + "grad_norm": 0.83984375, + "learning_rate": 0.00016376330927256682, + "loss": 2.4019, "step": 906 }, { - "epoch": 0.29047237790232183, - "grad_norm": 0.6171875, - "learning_rate": 4.4683234127555944e-05, - "loss": 2.3208, + "epoch": 0.35239552220684295, + "grad_norm": 0.765625, + "learning_rate": 0.00016365871084402108, + "loss": 2.4112, "step": 907 }, { - "epoch": 0.29079263410728584, - "grad_norm": 0.42578125, - "learning_rate": 4.46659835302746e-05, - "loss": 2.0736, + "epoch": 0.35278405089725845, + "grad_norm": 0.80078125, + "learning_rate": 0.00016355399518095052, + "loss": 2.3778, "step": 908 }, { - "epoch": 0.2911128903122498, - "grad_norm": 0.515625, - "learning_rate": 4.4648708334308156e-05, - "loss": 2.6272, + "epoch": 0.35317257958767395, + "grad_norm": 0.8203125, + "learning_rate": 0.0001634491624762006, + "loss": 2.4154, "step": 909 }, { - "epoch": 0.29143314651721375, - "grad_norm": 0.466796875, - "learning_rate": 4.463140856126485e-05, - "loss": 2.2224, + "epoch": 0.3535611082780894, + "grad_norm": 0.8828125, + "learning_rate": 0.0001633442129228322, + "loss": 2.4527, "step": 910 }, { - "epoch": 0.29175340272217776, - "grad_norm": 0.45703125, - "learning_rate": 4.461408423278365e-05, - "loss": 2.1467, + "epoch": 0.3539496369685049, + "grad_norm": 0.72265625, + "learning_rate": 0.00016323914671412154, + "loss": 2.3572, "step": 911 }, { - "epoch": 0.2920736589271417, - "grad_norm": 0.74609375, - "learning_rate": 4.4596735370534246e-05, - "loss": 2.5051, + "epoch": 0.3543381656589204, + "grad_norm": 0.93359375, + "learning_rate": 0.0001631339640435596, + "loss": 2.4506, "step": 912 }, { - "epoch": 0.29239391513210566, - "grad_norm": 0.466796875, - "learning_rate": 4.457936199621701e-05, - "loss": 2.0413, + "epoch": 0.35472669434933585, + "grad_norm": 0.79296875, + "learning_rate": 0.00016302866510485182, + "loss": 2.3681, "step": 913 }, { - "epoch": 0.2927141713370697, - "grad_norm": 0.5546875, - "learning_rate": 4.456196413156297e-05, - "loss": 2.1757, + "epoch": 0.35511522303975135, + "grad_norm": 0.7890625, + "learning_rate": 0.00016292325009191784, + "loss": 2.4713, "step": 914 }, { - "epoch": 0.2930344275420336, - "grad_norm": 0.68359375, - "learning_rate": 4.45445417983338e-05, - "loss": 2.51, + "epoch": 0.35550375173016685, + "grad_norm": 0.8125, + "learning_rate": 0.00016281771919889098, + "loss": 2.4329, "step": 915 }, { - "epoch": 0.2933546837469976, - "grad_norm": 0.64453125, - "learning_rate": 4.4527095018321786e-05, - "loss": 2.3372, + "epoch": 0.3558922804205823, + "grad_norm": 0.93359375, + "learning_rate": 0.000162712072620118, + "loss": 2.4268, "step": 916 }, { - "epoch": 0.2936749399519616, - "grad_norm": 0.82421875, - "learning_rate": 4.4509623813349755e-05, - "loss": 2.4337, + "epoch": 0.3562808091109978, + "grad_norm": 0.828125, + "learning_rate": 0.0001626063105501587, + "loss": 2.4348, "step": 917 }, { - "epoch": 0.29399519615692554, - "grad_norm": 1.109375, - "learning_rate": 4.4492128205271134e-05, - "loss": 2.0431, + "epoch": 0.3566693378014133, + "grad_norm": 0.74609375, + "learning_rate": 0.00016250043318378563, + "loss": 2.3795, "step": 918 }, { - "epoch": 0.2943154523618895, - "grad_norm": 0.51953125, - "learning_rate": 4.4474608215969825e-05, - "loss": 2.0454, + "epoch": 0.35705786649182875, + "grad_norm": 0.7890625, + "learning_rate": 0.00016239444071598354, + "loss": 2.4631, "step": 919 }, { - "epoch": 0.2946357085668535, - "grad_norm": 0.60546875, - "learning_rate": 4.445706386736027e-05, - "loss": 1.6327, + "epoch": 0.35744639518224425, + "grad_norm": 0.76171875, + "learning_rate": 0.00016228833334194927, + "loss": 2.3645, "step": 920 }, { - "epoch": 0.29495596477181746, - "grad_norm": 0.57421875, - "learning_rate": 4.4439495181387355e-05, - "loss": 2.1912, + "epoch": 0.35783492387265975, + "grad_norm": 0.734375, + "learning_rate": 0.00016218211125709124, + "loss": 2.4475, "step": 921 }, { - "epoch": 0.2952762209767814, - "grad_norm": 1.4609375, - "learning_rate": 4.442190218002643e-05, - "loss": 2.5112, + "epoch": 0.3582234525630752, + "grad_norm": 0.76171875, + "learning_rate": 0.00016207577465702908, + "loss": 2.4651, "step": 922 }, { - "epoch": 0.2955964771817454, - "grad_norm": 0.64453125, - "learning_rate": 4.440428488528322e-05, - "loss": 2.2843, + "epoch": 0.3586119812534907, + "grad_norm": 0.74609375, + "learning_rate": 0.00016196932373759338, + "loss": 2.416, "step": 923 }, { - "epoch": 0.2959167333867094, - "grad_norm": 0.408203125, - "learning_rate": 4.4386643319193874e-05, - "loss": 2.0363, + "epoch": 0.35900050994390614, + "grad_norm": 0.8515625, + "learning_rate": 0.0001618627586948252, + "loss": 2.3977, "step": 924 }, { - "epoch": 0.2962369895916733, - "grad_norm": 0.515625, - "learning_rate": 4.436897750382489e-05, - "loss": 2.4265, + "epoch": 0.35938903863432164, + "grad_norm": 0.76171875, + "learning_rate": 0.00016175607972497585, + "loss": 2.3937, "step": 925 }, { - "epoch": 0.29655724579663734, - "grad_norm": 0.58984375, - "learning_rate": 4.435128746127309e-05, - "loss": 2.0594, + "epoch": 0.35977756732473715, + "grad_norm": 0.73046875, + "learning_rate": 0.00016164928702450633, + "loss": 2.3936, "step": 926 }, { - "epoch": 0.2968775020016013, - "grad_norm": 0.482421875, - "learning_rate": 4.4333573213665604e-05, - "loss": 2.3537, + "epoch": 0.3601660960151526, + "grad_norm": 0.74609375, + "learning_rate": 0.00016154238079008719, + "loss": 2.3655, "step": 927 }, { - "epoch": 0.29719775820656524, - "grad_norm": 0.52734375, - "learning_rate": 4.431583478315984e-05, - "loss": 2.7564, + "epoch": 0.3605546247055681, + "grad_norm": 0.7734375, + "learning_rate": 0.00016143536121859803, + "loss": 2.4051, "step": 928 }, { - "epoch": 0.2975180144115292, - "grad_norm": 0.49609375, - "learning_rate": 4.429807219194345e-05, - "loss": 2.0438, + "epoch": 0.3609431533959836, + "grad_norm": 0.7734375, + "learning_rate": 0.0001613282285071272, + "loss": 2.3786, "step": 929 }, { - "epoch": 0.2978382706164932, - "grad_norm": 0.6015625, - "learning_rate": 4.428028546223432e-05, - "loss": 2.0035, + "epoch": 0.36133168208639904, + "grad_norm": 0.79296875, + "learning_rate": 0.00016122098285297132, + "loss": 2.4234, "step": 930 }, { - "epoch": 0.29815852682145716, - "grad_norm": 0.87109375, - "learning_rate": 4.4262474616280506e-05, - "loss": 1.9675, + "epoch": 0.36172021077681454, + "grad_norm": 0.8125, + "learning_rate": 0.00016111362445363512, + "loss": 2.3992, "step": 931 }, { - "epoch": 0.2984787830264211, - "grad_norm": 0.74609375, - "learning_rate": 4.4244639676360253e-05, - "loss": 2.2379, + "epoch": 0.36210873946723005, + "grad_norm": 0.71484375, + "learning_rate": 0.0001610061535068309, + "loss": 2.3751, "step": 932 }, { - "epoch": 0.2987990392313851, - "grad_norm": 1.8359375, - "learning_rate": 4.422678066478192e-05, - "loss": 1.8438, + "epoch": 0.3624972681576455, + "grad_norm": 0.8046875, + "learning_rate": 0.00016089857021047823, + "loss": 2.4276, "step": 933 }, { - "epoch": 0.2991192954363491, - "grad_norm": 0.4765625, - "learning_rate": 4.4208897603884016e-05, - "loss": 2.166, + "epoch": 0.362885796848061, + "grad_norm": 0.82421875, + "learning_rate": 0.0001607908747627036, + "loss": 2.3446, "step": 934 }, { - "epoch": 0.29943955164131303, - "grad_norm": 0.4609375, - "learning_rate": 4.419099051603507e-05, - "loss": 2.4782, + "epoch": 0.3632743255384765, + "grad_norm": 0.95703125, + "learning_rate": 0.00016068306736184004, + "loss": 2.3987, "step": 935 }, { - "epoch": 0.29975980784627704, - "grad_norm": 0.37890625, - "learning_rate": 4.417305942363371e-05, - "loss": 1.9755, + "epoch": 0.36366285422889194, + "grad_norm": 0.76953125, + "learning_rate": 0.00016057514820642677, + "loss": 2.4452, "step": 936 }, { - "epoch": 0.300080064051241, - "grad_norm": 0.5625, - "learning_rate": 4.415510434910859e-05, - "loss": 2.6677, + "epoch": 0.36405138291930744, + "grad_norm": 0.8515625, + "learning_rate": 0.00016046711749520876, + "loss": 2.4444, "step": 937 }, { - "epoch": 0.30040032025620494, - "grad_norm": 0.494140625, - "learning_rate": 4.413712531491833e-05, - "loss": 2.0538, + "epoch": 0.36443991160972294, + "grad_norm": 0.78515625, + "learning_rate": 0.00016035897542713648, + "loss": 2.3775, "step": 938 }, { - "epoch": 0.30072057646116895, - "grad_norm": 0.6328125, - "learning_rate": 4.411912234355156e-05, - "loss": 2.4744, + "epoch": 0.3648284403001384, + "grad_norm": 0.8671875, + "learning_rate": 0.00016025072220136543, + "loss": 2.4025, "step": 939 }, { - "epoch": 0.3010408326661329, - "grad_norm": 0.8203125, - "learning_rate": 4.41010954575268e-05, - "loss": 2.4799, + "epoch": 0.3652169689905539, + "grad_norm": 0.81640625, + "learning_rate": 0.00016014235801725587, + "loss": 2.4327, "step": 940 }, { - "epoch": 0.30136108887109686, - "grad_norm": 0.4609375, - "learning_rate": 4.408304467939254e-05, - "loss": 2.5785, + "epoch": 0.3656054976809694, + "grad_norm": 0.82421875, + "learning_rate": 0.00016003388307437238, + "loss": 2.4564, "step": 941 }, { - "epoch": 0.30168134507606087, - "grad_norm": 0.65234375, - "learning_rate": 4.4064970031727114e-05, - "loss": 2.2337, + "epoch": 0.36599402637138484, + "grad_norm": 0.84765625, + "learning_rate": 0.00015992529757248347, + "loss": 2.4282, "step": 942 }, { - "epoch": 0.3020016012810248, - "grad_norm": 0.5703125, - "learning_rate": 4.4046871537138734e-05, - "loss": 2.5195, + "epoch": 0.36638255506180034, + "grad_norm": 0.7578125, + "learning_rate": 0.00015981660171156134, + "loss": 2.3523, "step": 943 }, { - "epoch": 0.3023218574859888, - "grad_norm": 0.62109375, - "learning_rate": 4.402874921826542e-05, - "loss": 2.3741, + "epoch": 0.36677108375221584, + "grad_norm": 0.78125, + "learning_rate": 0.00015970779569178138, + "loss": 2.4013, "step": 944 }, { - "epoch": 0.3026421136909528, - "grad_norm": 0.5234375, - "learning_rate": 4.4010603097775015e-05, - "loss": 2.4184, + "epoch": 0.3671596124426313, + "grad_norm": 0.78515625, + "learning_rate": 0.00015959887971352184, + "loss": 2.3586, "step": 945 }, { - "epoch": 0.30296236989591674, - "grad_norm": 0.267578125, - "learning_rate": 4.399243319836512e-05, - "loss": 2.1592, + "epoch": 0.3675481411330468, + "grad_norm": 0.84375, + "learning_rate": 0.00015948985397736348, + "loss": 2.4387, "step": 946 }, { - "epoch": 0.3032826261008807, - "grad_norm": 0.419921875, - "learning_rate": 4.397423954276308e-05, - "loss": 1.8628, + "epoch": 0.3679366698234623, + "grad_norm": 0.8125, + "learning_rate": 0.00015938071868408922, + "loss": 2.3555, "step": 947 }, { - "epoch": 0.3036028823058447, - "grad_norm": 0.439453125, - "learning_rate": 4.395602215372596e-05, - "loss": 1.8973, + "epoch": 0.36832519851387774, + "grad_norm": 0.75, + "learning_rate": 0.00015927147403468369, + "loss": 2.3974, "step": 948 }, { - "epoch": 0.30392313851080865, - "grad_norm": 0.5546875, - "learning_rate": 4.3937781054040505e-05, - "loss": 2.1797, + "epoch": 0.36871372720429324, + "grad_norm": 0.7890625, + "learning_rate": 0.00015916212023033298, + "loss": 2.4057, "step": 949 }, { - "epoch": 0.3042433947157726, - "grad_norm": 2.40625, - "learning_rate": 4.391951626652312e-05, - "loss": 2.0715, + "epoch": 0.36910225589470874, + "grad_norm": 0.75, + "learning_rate": 0.00015905265747242413, + "loss": 2.4402, "step": 950 }, { - "epoch": 0.30456365092073656, - "grad_norm": 0.40234375, - "learning_rate": 4.3901227814019845e-05, - "loss": 1.859, + "epoch": 0.3694907845851242, + "grad_norm": 0.75390625, + "learning_rate": 0.00015894308596254485, + "loss": 2.4567, "step": 951 }, { - "epoch": 0.30488390712570057, - "grad_norm": 0.40234375, - "learning_rate": 4.388291571940631e-05, - "loss": 2.4701, + "epoch": 0.3698793132755397, + "grad_norm": 0.80078125, + "learning_rate": 0.00015883340590248314, + "loss": 2.4268, "step": 952 }, { - "epoch": 0.3052041633306645, - "grad_norm": 0.53125, - "learning_rate": 4.386458000558773e-05, - "loss": 2.0972, + "epoch": 0.3702678419659552, + "grad_norm": 0.80078125, + "learning_rate": 0.00015872361749422694, + "loss": 2.379, "step": 953 }, { - "epoch": 0.3055244195356285, - "grad_norm": 8.5, - "learning_rate": 4.384622069549886e-05, - "loss": 1.9298, + "epoch": 0.37065637065637064, + "grad_norm": 0.75390625, + "learning_rate": 0.0001586137209399637, + "loss": 2.4159, "step": 954 }, { - "epoch": 0.3058446757405925, - "grad_norm": 0.5859375, - "learning_rate": 4.382783781210395e-05, - "loss": 2.1858, + "epoch": 0.37104489934678614, + "grad_norm": 0.80859375, + "learning_rate": 0.00015850371644207995, + "loss": 2.4021, "step": 955 }, { - "epoch": 0.30616493194555644, - "grad_norm": 0.453125, - "learning_rate": 4.3809431378396776e-05, - "loss": 2.1796, + "epoch": 0.37143342803720164, + "grad_norm": 0.734375, + "learning_rate": 0.00015839360420316116, + "loss": 2.4081, "step": 956 }, { - "epoch": 0.3064851881505204, - "grad_norm": 0.50390625, - "learning_rate": 4.379100141740053e-05, - "loss": 2.071, + "epoch": 0.3718219567276171, + "grad_norm": 0.78515625, + "learning_rate": 0.00015828338442599112, + "loss": 2.4259, "step": 957 }, { - "epoch": 0.3068054443554844, - "grad_norm": 1.8359375, - "learning_rate": 4.377254795216785e-05, - "loss": 2.5639, + "epoch": 0.3722104854180326, + "grad_norm": 0.76953125, + "learning_rate": 0.00015817305731355167, + "loss": 2.4069, "step": 958 }, { - "epoch": 0.30712570056044836, - "grad_norm": 0.859375, - "learning_rate": 4.3754071005780785e-05, - "loss": 2.4597, + "epoch": 0.3725990141084481, + "grad_norm": 0.77734375, + "learning_rate": 0.00015806262306902234, + "loss": 2.4049, "step": 959 }, { - "epoch": 0.3074459567654123, - "grad_norm": 0.8984375, - "learning_rate": 4.3735570601350736e-05, - "loss": 2.6462, + "epoch": 0.37298754279886354, + "grad_norm": 0.8125, + "learning_rate": 0.00015795208189577995, + "loss": 2.412, "step": 960 }, { - "epoch": 0.3077662129703763, - "grad_norm": 0.5390625, - "learning_rate": 4.3717046762018454e-05, - "loss": 2.2894, + "epoch": 0.37337607148927904, + "grad_norm": 0.796875, + "learning_rate": 0.00015784143399739822, + "loss": 2.4012, "step": 961 }, { - "epoch": 0.30808646917534027, - "grad_norm": 0.486328125, - "learning_rate": 4.369849951095401e-05, - "loss": 1.6288, + "epoch": 0.37376460017969454, + "grad_norm": 0.8515625, + "learning_rate": 0.00015773067957764752, + "loss": 2.3994, "step": 962 }, { - "epoch": 0.3084067253803042, - "grad_norm": 0.298828125, - "learning_rate": 4.3679928871356743e-05, - "loss": 1.4223, + "epoch": 0.37415312887011, + "grad_norm": 0.80078125, + "learning_rate": 0.0001576198188404942, + "loss": 2.4166, "step": 963 }, { - "epoch": 0.30872698158526823, - "grad_norm": 0.5390625, - "learning_rate": 4.366133486645526e-05, - "loss": 2.0063, + "epoch": 0.3745416575605255, + "grad_norm": 0.7421875, + "learning_rate": 0.0001575088519901006, + "loss": 2.3468, "step": 964 }, { - "epoch": 0.3090472377902322, - "grad_norm": 0.4921875, - "learning_rate": 4.364271751950739e-05, - "loss": 2.1687, + "epoch": 0.374930186250941, + "grad_norm": 0.76953125, + "learning_rate": 0.0001573977792308243, + "loss": 2.4512, "step": 965 }, { - "epoch": 0.30936749399519614, - "grad_norm": 0.6015625, - "learning_rate": 4.362407685380015e-05, - "loss": 2.4734, + "epoch": 0.37531871494135643, + "grad_norm": 0.7734375, + "learning_rate": 0.00015728660076721812, + "loss": 2.3884, "step": 966 }, { - "epoch": 0.30968775020016015, - "grad_norm": 0.51171875, - "learning_rate": 4.3605412892649744e-05, - "loss": 2.215, + "epoch": 0.37570724363177194, + "grad_norm": 0.83203125, + "learning_rate": 0.00015717531680402934, + "loss": 2.4164, "step": 967 }, { - "epoch": 0.3100080064051241, - "grad_norm": 1.5078125, - "learning_rate": 4.3586725659401504e-05, - "loss": 2.1618, + "epoch": 0.37609577232218744, + "grad_norm": 0.78515625, + "learning_rate": 0.00015706392754619974, + "loss": 2.4216, "step": 968 }, { - "epoch": 0.31032826261008806, - "grad_norm": 0.4296875, - "learning_rate": 4.356801517742986e-05, - "loss": 2.6403, + "epoch": 0.3764843010126029, + "grad_norm": 0.91015625, + "learning_rate": 0.00015695243319886483, + "loss": 2.3632, "step": 969 }, { - "epoch": 0.31064851881505207, - "grad_norm": 0.51171875, - "learning_rate": 4.3549281470138334e-05, - "loss": 2.0227, + "epoch": 0.3768728297030184, + "grad_norm": 0.828125, + "learning_rate": 0.00015684083396735372, + "loss": 2.3983, "step": 970 }, { - "epoch": 0.310968775020016, - "grad_norm": 0.7578125, - "learning_rate": 4.353052456095952e-05, - "loss": 2.4126, + "epoch": 0.3772613583934339, + "grad_norm": 0.796875, + "learning_rate": 0.0001567291300571887, + "loss": 2.4777, "step": 971 }, { - "epoch": 0.31128903122498, - "grad_norm": 0.609375, - "learning_rate": 4.351174447335498e-05, - "loss": 2.1611, + "epoch": 0.37764988708384933, + "grad_norm": 0.8203125, + "learning_rate": 0.00015661732167408484, + "loss": 2.431, "step": 972 }, { - "epoch": 0.311609287429944, - "grad_norm": 0.59375, - "learning_rate": 4.349294123081531e-05, - "loss": 2.2805, + "epoch": 0.37803841577426484, + "grad_norm": 0.8984375, + "learning_rate": 0.00015650540902394954, + "loss": 2.486, "step": 973 }, { - "epoch": 0.31192954363490794, - "grad_norm": 0.56640625, - "learning_rate": 4.347411485686006e-05, - "loss": 2.2679, + "epoch": 0.37842694446468034, + "grad_norm": 0.78125, + "learning_rate": 0.00015639339231288233, + "loss": 2.4245, "step": 974 }, { - "epoch": 0.3122497998398719, - "grad_norm": 0.7421875, - "learning_rate": 4.345526537503771e-05, - "loss": 2.2008, + "epoch": 0.3788154731550958, + "grad_norm": 0.7578125, + "learning_rate": 0.00015628127174717427, + "loss": 2.3979, "step": 975 }, { - "epoch": 0.31257005604483584, - "grad_norm": 2.0, - "learning_rate": 4.3436392808925654e-05, - "loss": 1.8056, + "epoch": 0.3792040018455113, + "grad_norm": 0.75, + "learning_rate": 0.00015616904753330775, + "loss": 2.3735, "step": 976 }, { - "epoch": 0.31289031224979985, - "grad_norm": 0.5234375, - "learning_rate": 4.341749718213014e-05, - "loss": 2.3376, + "epoch": 0.3795925305359268, + "grad_norm": 0.77734375, + "learning_rate": 0.00015605671987795595, + "loss": 2.4248, "step": 977 }, { - "epoch": 0.3132105684547638, - "grad_norm": 0.91015625, - "learning_rate": 4.339857851828628e-05, + "epoch": 0.37998105922634223, + "grad_norm": 0.7578125, + "learning_rate": 0.00015594428898798272, "loss": 2.4653, "step": 978 }, { - "epoch": 0.31353082465972776, - "grad_norm": 0.828125, - "learning_rate": 4.337963684105799e-05, - "loss": 2.2227, + "epoch": 0.38036958791675773, + "grad_norm": 0.77734375, + "learning_rate": 0.00015583175507044185, + "loss": 2.4554, "step": 979 }, { - "epoch": 0.31385108086469177, - "grad_norm": 0.375, - "learning_rate": 4.336067217413797e-05, - "loss": 2.2355, + "epoch": 0.38075811660717324, + "grad_norm": 0.8671875, + "learning_rate": 0.00015571911833257696, + "loss": 2.4347, "step": 980 }, { - "epoch": 0.3141713370696557, - "grad_norm": 0.490234375, - "learning_rate": 4.334168454124769e-05, - "loss": 2.4895, + "epoch": 0.3811466452975887, + "grad_norm": 0.78515625, + "learning_rate": 0.000155606378981821, + "loss": 2.4075, "step": 981 }, { - "epoch": 0.3144915932746197, - "grad_norm": 0.43359375, - "learning_rate": 4.332267396613734e-05, - "loss": 1.9417, + "epoch": 0.3815351739880042, + "grad_norm": 0.6953125, + "learning_rate": 0.00015549353722579592, + "loss": 2.3812, "step": 982 }, { - "epoch": 0.3148118494795837, - "grad_norm": 0.53125, - "learning_rate": 4.330364047258579e-05, - "loss": 2.6751, + "epoch": 0.3819237026784197, + "grad_norm": 0.6953125, + "learning_rate": 0.0001553805932723122, + "loss": 2.2825, "step": 983 }, { - "epoch": 0.31513210568454764, - "grad_norm": 0.369140625, - "learning_rate": 4.3284584084400604e-05, - "loss": 1.9346, + "epoch": 0.38231223136883513, + "grad_norm": 0.78515625, + "learning_rate": 0.00015526754732936862, + "loss": 2.4434, "step": 984 }, { - "epoch": 0.3154523618895116, - "grad_norm": 0.46875, - "learning_rate": 4.3265504825417966e-05, - "loss": 2.1739, + "epoch": 0.38270076005925063, + "grad_norm": 0.7578125, + "learning_rate": 0.00015515439960515166, + "loss": 2.3885, "step": 985 }, { - "epoch": 0.3157726180944756, - "grad_norm": 0.388671875, - "learning_rate": 4.3246402719502676e-05, - "loss": 1.7595, + "epoch": 0.38308928874966613, + "grad_norm": 0.7578125, + "learning_rate": 0.00015504115030803535, + "loss": 2.3675, "step": 986 }, { - "epoch": 0.31609287429943955, - "grad_norm": 0.5078125, - "learning_rate": 4.3227277790548104e-05, - "loss": 2.2438, + "epoch": 0.3834778174400816, + "grad_norm": 0.75, + "learning_rate": 0.00015492779964658074, + "loss": 2.3827, "step": 987 }, { - "epoch": 0.3164131305044035, - "grad_norm": 0.546875, - "learning_rate": 4.3208130062476174e-05, - "loss": 2.4872, + "epoch": 0.3838663461304971, + "grad_norm": 0.859375, + "learning_rate": 0.00015481434782953555, + "loss": 2.4087, "step": 988 }, { - "epoch": 0.3167333867093675, - "grad_norm": 0.46875, - "learning_rate": 4.318895955923731e-05, - "loss": 2.4345, + "epoch": 0.3842548748209126, + "grad_norm": 0.86328125, + "learning_rate": 0.00015470079506583377, + "loss": 2.4245, "step": 989 }, { - "epoch": 0.31705364291433147, - "grad_norm": 0.447265625, - "learning_rate": 4.316976630481046e-05, - "loss": 2.2613, + "epoch": 0.38464340351132803, + "grad_norm": 0.7890625, + "learning_rate": 0.00015458714156459536, + "loss": 2.376, "step": 990 }, { - "epoch": 0.3173738991192954, - "grad_norm": 0.46484375, - "learning_rate": 4.3150550323203e-05, - "loss": 2.1071, + "epoch": 0.38503193220174353, + "grad_norm": 0.90625, + "learning_rate": 0.00015447338753512573, + "loss": 2.3831, "step": 991 }, { - "epoch": 0.31769415532425943, - "grad_norm": 0.5234375, - "learning_rate": 4.313131163845073e-05, - "loss": 2.1463, + "epoch": 0.385420460892159, + "grad_norm": 0.92578125, + "learning_rate": 0.00015435953318691543, + "loss": 2.4737, "step": 992 }, { - "epoch": 0.3180144115292234, - "grad_norm": 0.98828125, - "learning_rate": 4.311205027461789e-05, - "loss": 1.4839, + "epoch": 0.3858089895825745, + "grad_norm": 0.796875, + "learning_rate": 0.00015424557872963982, + "loss": 2.4105, "step": 993 }, { - "epoch": 0.31833466773418734, - "grad_norm": 0.640625, - "learning_rate": 4.309276625579703e-05, - "loss": 2.4197, + "epoch": 0.38619751827299, + "grad_norm": 0.89453125, + "learning_rate": 0.00015413152437315856, + "loss": 2.4475, "step": 994 }, { - "epoch": 0.31865492393915135, - "grad_norm": 0.35546875, - "learning_rate": 4.307345960610906e-05, - "loss": 2.0932, + "epoch": 0.3865860469634054, + "grad_norm": 0.78125, + "learning_rate": 0.00015401737032751532, + "loss": 2.455, "step": 995 }, { - "epoch": 0.3189751801441153, - "grad_norm": 0.5703125, - "learning_rate": 4.305413034970322e-05, - "loss": 2.1792, + "epoch": 0.38697457565382093, + "grad_norm": 0.734375, + "learning_rate": 0.00015390311680293727, + "loss": 2.3793, "step": 996 }, { - "epoch": 0.31929543634907925, - "grad_norm": 0.421875, - "learning_rate": 4.3034778510757e-05, - "loss": 2.5835, + "epoch": 0.38736310434423643, + "grad_norm": 0.78515625, + "learning_rate": 0.00015378876400983494, + "loss": 2.406, "step": 997 }, { - "epoch": 0.3196156925540432, - "grad_norm": 0.5078125, - "learning_rate": 4.3015404113476156e-05, - "loss": 1.9542, + "epoch": 0.3877516330346519, + "grad_norm": 0.8203125, + "learning_rate": 0.00015367431215880157, + "loss": 2.4224, "step": 998 }, { - "epoch": 0.3199359487590072, - "grad_norm": 0.384765625, - "learning_rate": 4.2996007182094634e-05, - "loss": 1.7533, + "epoch": 0.3881401617250674, + "grad_norm": 1.015625, + "learning_rate": 0.00015355976146061285, + "loss": 2.4179, "step": 999 }, { - "epoch": 0.32025620496397117, - "grad_norm": 0.66796875, - "learning_rate": 4.297658774087459e-05, - "loss": 2.4536, + "epoch": 0.3885286904154829, + "grad_norm": 0.80859375, + "learning_rate": 0.00015344511212622648, + "loss": 2.4217, "step": 1000 }, { - "epoch": 0.3205764611689351, - "grad_norm": 0.796875, - "learning_rate": 4.295714581410633e-05, - "loss": 2.0973, + "epoch": 0.3889172191058983, + "grad_norm": 0.80859375, + "learning_rate": 0.00015333036436678183, + "loss": 2.4036, "step": 1001 }, { - "epoch": 0.32089671737389913, - "grad_norm": 0.419921875, - "learning_rate": 4.293768142610828e-05, - "loss": 2.5078, + "epoch": 0.38930574779631383, + "grad_norm": 0.78125, + "learning_rate": 0.00015321551839359953, + "loss": 2.472, "step": 1002 }, { - "epoch": 0.3212169735788631, - "grad_norm": 0.7734375, - "learning_rate": 4.291819460122697e-05, - "loss": 1.7711, + "epoch": 0.38969427648672933, + "grad_norm": 0.87890625, + "learning_rate": 0.00015310057441818113, + "loss": 2.4167, "step": 1003 }, { - "epoch": 0.32153722978382704, - "grad_norm": 0.3828125, - "learning_rate": 4.2898685363836974e-05, - "loss": 2.1171, + "epoch": 0.3900828051771448, + "grad_norm": 0.7890625, + "learning_rate": 0.00015298553265220854, + "loss": 2.4069, "step": 1004 }, { - "epoch": 0.32185748598879105, - "grad_norm": 0.5234375, - "learning_rate": 4.287915373834094e-05, - "loss": 2.2499, + "epoch": 0.3904713338675603, + "grad_norm": 0.76953125, + "learning_rate": 0.0001528703933075439, + "loss": 2.3354, "step": 1005 }, { - "epoch": 0.322177742193755, - "grad_norm": 0.51171875, - "learning_rate": 4.2859599749169474e-05, - "loss": 1.9514, + "epoch": 0.3908598625579758, + "grad_norm": 0.74609375, + "learning_rate": 0.00015275515659622898, + "loss": 2.3972, "step": 1006 }, { - "epoch": 0.32249799839871895, - "grad_norm": 0.44140625, - "learning_rate": 4.284002342078119e-05, - "loss": 2.2806, + "epoch": 0.3912483912483912, + "grad_norm": 0.79296875, + "learning_rate": 0.00015263982273048488, + "loss": 2.4468, "step": 1007 }, { - "epoch": 0.32281825460368296, - "grad_norm": 0.484375, - "learning_rate": 4.2820424777662616e-05, - "loss": 1.9288, + "epoch": 0.3916369199388067, + "grad_norm": 0.7890625, + "learning_rate": 0.00015252439192271156, + "loss": 2.4281, "step": 1008 }, { - "epoch": 0.3231385108086469, - "grad_norm": 0.55859375, - "learning_rate": 4.280080384432821e-05, - "loss": 2.4497, + "epoch": 0.39202544862922223, + "grad_norm": 0.78125, + "learning_rate": 0.0001524088643854876, + "loss": 2.3852, "step": 1009 }, { - "epoch": 0.32345876701361087, - "grad_norm": 0.53125, - "learning_rate": 4.278116064532032e-05, - "loss": 2.273, + "epoch": 0.3924139773196377, + "grad_norm": 0.97265625, + "learning_rate": 0.00015229324033156968, + "loss": 2.4179, "step": 1010 }, { - "epoch": 0.3237790232185749, - "grad_norm": 0.64453125, - "learning_rate": 4.2761495205209114e-05, - "loss": 2.2382, + "epoch": 0.3928025060100532, + "grad_norm": 0.83984375, + "learning_rate": 0.00015217751997389223, + "loss": 2.4206, "step": 1011 }, { - "epoch": 0.32409927942353883, - "grad_norm": 0.4921875, - "learning_rate": 4.274180754859261e-05, - "loss": 2.0357, + "epoch": 0.3931910347004687, + "grad_norm": 0.76953125, + "learning_rate": 0.00015206170352556703, + "loss": 2.3465, "step": 1012 }, { - "epoch": 0.3244195356285028, - "grad_norm": 0.6171875, - "learning_rate": 4.27220977000966e-05, - "loss": 1.655, + "epoch": 0.3935795633908841, + "grad_norm": 0.76171875, + "learning_rate": 0.0001519457911998828, + "loss": 2.3244, "step": 1013 }, { - "epoch": 0.3247397918334668, - "grad_norm": 0.49609375, - "learning_rate": 4.2702365684374626e-05, - "loss": 2.2382, + "epoch": 0.3939680920812996, + "grad_norm": 0.796875, + "learning_rate": 0.00015182978321030482, + "loss": 2.3715, "step": 1014 }, { - "epoch": 0.32506004803843075, - "grad_norm": 0.322265625, - "learning_rate": 4.2682611526107986e-05, - "loss": 1.7932, + "epoch": 0.3943566207717151, + "grad_norm": 0.77734375, + "learning_rate": 0.00015171367977047465, + "loss": 2.396, "step": 1015 }, { - "epoch": 0.3253803042433947, - "grad_norm": 0.59375, - "learning_rate": 4.266283525000564e-05, - "loss": 2.5412, + "epoch": 0.3947451494621306, + "grad_norm": 0.8828125, + "learning_rate": 0.00015159748109420956, + "loss": 2.4217, "step": 1016 }, { - "epoch": 0.3257005604483587, - "grad_norm": 0.6953125, - "learning_rate": 4.2643036880804236e-05, - "loss": 2.1839, + "epoch": 0.3951336781525461, + "grad_norm": 0.7265625, + "learning_rate": 0.00015148118739550213, + "loss": 2.4046, "step": 1017 }, { - "epoch": 0.32602081665332266, - "grad_norm": 0.5078125, - "learning_rate": 4.2623216443268055e-05, - "loss": 2.3921, + "epoch": 0.3955222068429616, + "grad_norm": 0.76171875, + "learning_rate": 0.00015136479888852006, + "loss": 2.3725, "step": 1018 }, { - "epoch": 0.3263410728582866, - "grad_norm": 0.53515625, - "learning_rate": 4.2603373962188966e-05, - "loss": 2.1971, + "epoch": 0.395910735533377, + "grad_norm": 0.85546875, + "learning_rate": 0.00015124831578760558, + "loss": 2.3309, "step": 1019 }, { - "epoch": 0.3266613290632506, - "grad_norm": 0.7890625, - "learning_rate": 4.258350946238643e-05, - "loss": 2.4754, + "epoch": 0.3962992642237925, + "grad_norm": 0.8125, + "learning_rate": 0.00015113173830727514, + "loss": 2.4652, "step": 1020 }, { - "epoch": 0.3269815852682146, - "grad_norm": 0.60546875, - "learning_rate": 4.256362296870744e-05, - "loss": 2.1899, + "epoch": 0.396687792914208, + "grad_norm": 0.74609375, + "learning_rate": 0.00015101506666221904, + "loss": 2.3596, "step": 1021 }, { - "epoch": 0.32730184147317853, - "grad_norm": 0.47265625, - "learning_rate": 4.25437145060265e-05, - "loss": 2.5574, + "epoch": 0.3970763216046235, + "grad_norm": 0.73046875, + "learning_rate": 0.00015089830106730095, + "loss": 2.311, "step": 1022 }, { - "epoch": 0.3276220976781425, - "grad_norm": 0.55859375, - "learning_rate": 4.252378409924559e-05, - "loss": 2.1373, + "epoch": 0.397464850295039, + "grad_norm": 0.7578125, + "learning_rate": 0.00015078144173755754, + "loss": 2.3798, "step": 1023 }, { - "epoch": 0.3279423538831065, - "grad_norm": 0.439453125, - "learning_rate": 4.2503831773294144e-05, - "loss": 2.4083, + "epoch": 0.3978533789854545, + "grad_norm": 0.734375, + "learning_rate": 0.00015066448888819816, + "loss": 2.3415, "step": 1024 }, { - "epoch": 0.32826261008807045, - "grad_norm": 0.466796875, - "learning_rate": 4.248385755312901e-05, - "loss": 1.7979, + "epoch": 0.3982419076758699, + "grad_norm": 0.78125, + "learning_rate": 0.00015054744273460436, + "loss": 2.4462, "step": 1025 }, { - "epoch": 0.3285828662930344, - "grad_norm": 0.546875, - "learning_rate": 4.246386146373444e-05, - "loss": 2.0333, + "epoch": 0.3986304363662854, + "grad_norm": 0.75390625, + "learning_rate": 0.00015043030349232947, + "loss": 2.3352, "step": 1026 }, { - "epoch": 0.3289031224979984, - "grad_norm": 0.404296875, - "learning_rate": 4.244384353012199e-05, - "loss": 1.9149, + "epoch": 0.3990189650567009, + "grad_norm": 0.75390625, + "learning_rate": 0.00015031307137709833, + "loss": 2.3445, "step": 1027 }, { - "epoch": 0.32922337870296237, - "grad_norm": 0.796875, - "learning_rate": 4.2423803777330606e-05, - "loss": 2.0417, + "epoch": 0.39940749374711637, + "grad_norm": 0.73828125, + "learning_rate": 0.00015019574660480684, + "loss": 2.342, "step": 1028 }, { - "epoch": 0.3295436349079263, - "grad_norm": 0.431640625, - "learning_rate": 4.240374223042647e-05, - "loss": 1.9744, + "epoch": 0.3997960224375319, + "grad_norm": 0.8203125, + "learning_rate": 0.00015007832939152143, + "loss": 2.4521, "step": 1029 }, { - "epoch": 0.32986389111289033, - "grad_norm": 0.484375, - "learning_rate": 4.2383658914503056e-05, - "loss": 2.2739, + "epoch": 0.4001845511279474, + "grad_norm": 0.78125, + "learning_rate": 0.00014996081995347885, + "loss": 2.3904, "step": 1030 }, { - "epoch": 0.3301841473178543, - "grad_norm": 0.30078125, - "learning_rate": 4.236355385468106e-05, - "loss": 1.5005, + "epoch": 0.4005730798183628, + "grad_norm": 0.7890625, + "learning_rate": 0.00014984321850708562, + "loss": 2.4335, "step": 1031 }, { - "epoch": 0.33050440352281824, - "grad_norm": 0.451171875, - "learning_rate": 4.234342707610837e-05, - "loss": 2.1598, + "epoch": 0.4009616085087783, + "grad_norm": 0.75390625, + "learning_rate": 0.00014972552526891782, + "loss": 2.4618, "step": 1032 }, { - "epoch": 0.33082465972778224, - "grad_norm": 0.99609375, - "learning_rate": 4.2323278603960046e-05, - "loss": 2.4382, + "epoch": 0.4013501371991938, + "grad_norm": 0.76171875, + "learning_rate": 0.00014960774045572045, + "loss": 2.4655, "step": 1033 }, { - "epoch": 0.3311449159327462, - "grad_norm": 0.380859375, - "learning_rate": 4.230310846343828e-05, - "loss": 1.963, + "epoch": 0.40173866588960927, + "grad_norm": 0.703125, + "learning_rate": 0.00014948986428440727, + "loss": 2.2844, "step": 1034 }, { - "epoch": 0.33146517213771015, - "grad_norm": 0.390625, - "learning_rate": 4.228291667977238e-05, - "loss": 1.9068, + "epoch": 0.40212719458002477, + "grad_norm": 0.734375, + "learning_rate": 0.0001493718969720602, + "loss": 2.3762, "step": 1035 }, { - "epoch": 0.33178542834267416, - "grad_norm": 0.42578125, - "learning_rate": 4.22627032782187e-05, - "loss": 1.4926, + "epoch": 0.4025157232704403, + "grad_norm": 0.78125, + "learning_rate": 0.00014925383873592895, + "loss": 2.4462, "step": 1036 }, { - "epoch": 0.3321056845476381, - "grad_norm": 0.5, - "learning_rate": 4.2242468284060644e-05, - "loss": 2.3178, + "epoch": 0.4029042519608557, + "grad_norm": 0.73828125, + "learning_rate": 0.0001491356897934309, + "loss": 2.4108, "step": 1037 }, { - "epoch": 0.33242594075260207, - "grad_norm": 0.41796875, - "learning_rate": 4.222221172260865e-05, - "loss": 2.6006, + "epoch": 0.4032927806512712, + "grad_norm": 0.79296875, + "learning_rate": 0.00014901745036215023, + "loss": 2.4457, "step": 1038 }, { - "epoch": 0.3327461969575661, - "grad_norm": 0.5390625, - "learning_rate": 4.2201933619200095e-05, - "loss": 2.1655, + "epoch": 0.4036813093416867, + "grad_norm": 0.74609375, + "learning_rate": 0.00014889912065983794, + "loss": 2.2958, "step": 1039 }, { - "epoch": 0.33306645316253003, - "grad_norm": 0.875, - "learning_rate": 4.218163399919933e-05, - "loss": 2.2719, + "epoch": 0.40406983803210217, + "grad_norm": 0.76171875, + "learning_rate": 0.0001487807009044112, + "loss": 2.3535, "step": 1040 }, { - "epoch": 0.333386709367494, - "grad_norm": 0.8203125, - "learning_rate": 4.216131288799761e-05, - "loss": 2.1145, + "epoch": 0.40445836672251767, + "grad_norm": 0.75390625, + "learning_rate": 0.00014866219131395297, + "loss": 2.403, "step": 1041 }, { - "epoch": 0.333706965572458, - "grad_norm": 0.498046875, - "learning_rate": 4.214097031101305e-05, - "loss": 1.9372, - "step": 1042 + "epoch": 0.40484689541293317, + "grad_norm": 0.8515625, + "learning_rate": 0.00014854359210671173, + "loss": 2.397, + "step": 1042 }, { - "epoch": 0.33402722177742195, - "grad_norm": 0.59765625, - "learning_rate": 4.212060629369065e-05, - "loss": 2.4681, + "epoch": 0.4052354241033486, + "grad_norm": 0.76953125, + "learning_rate": 0.00014842490350110103, + "loss": 2.4236, "step": 1043 }, { - "epoch": 0.3343474779823859, - "grad_norm": 0.37890625, - "learning_rate": 4.210022086150221e-05, - "loss": 1.957, + "epoch": 0.4056239527937641, + "grad_norm": 0.84375, + "learning_rate": 0.00014830612571569896, + "loss": 2.3644, "step": 1044 }, { - "epoch": 0.33466773418734985, - "grad_norm": 0.36328125, - "learning_rate": 4.207981403994632e-05, - "loss": 2.0795, + "epoch": 0.4060124814841796, + "grad_norm": 0.74609375, + "learning_rate": 0.00014818725896924786, + "loss": 2.4009, "step": 1045 }, { - "epoch": 0.33498799039231386, - "grad_norm": 0.52734375, - "learning_rate": 4.2059385854548316e-05, - "loss": 2.4652, + "epoch": 0.40640101017459507, + "grad_norm": 0.73046875, + "learning_rate": 0.000148068303480654, + "loss": 2.3757, "step": 1046 }, { - "epoch": 0.3353082465972778, - "grad_norm": 0.51171875, - "learning_rate": 4.203893633086027e-05, - "loss": 2.3613, + "epoch": 0.40678953886501057, + "grad_norm": 0.796875, + "learning_rate": 0.000147949259468987, + "loss": 2.3917, "step": 1047 }, { - "epoch": 0.33562850280224177, - "grad_norm": 0.67578125, - "learning_rate": 4.201846549446094e-05, - "loss": 2.4627, + "epoch": 0.40717806755542607, + "grad_norm": 0.8046875, + "learning_rate": 0.00014783012715347945, + "loss": 2.4178, "step": 1048 }, { - "epoch": 0.3359487590072058, - "grad_norm": 0.408203125, - "learning_rate": 4.1997973370955734e-05, - "loss": 2.232, + "epoch": 0.4075665962458415, + "grad_norm": 0.84765625, + "learning_rate": 0.00014771090675352665, + "loss": 2.412, "step": 1049 }, { - "epoch": 0.33626901521216973, - "grad_norm": 0.421875, - "learning_rate": 4.19774599859767e-05, - "loss": 2.6032, + "epoch": 0.407955124936257, + "grad_norm": 0.7578125, + "learning_rate": 0.00014759159848868613, + "loss": 2.4158, "step": 1050 }, { - "epoch": 0.3365892714171337, - "grad_norm": 0.455078125, - "learning_rate": 4.1956925365182466e-05, - "loss": 2.4148, + "epoch": 0.4083436536266725, + "grad_norm": 0.7578125, + "learning_rate": 0.00014747220257867717, + "loss": 2.3423, "step": 1051 }, { - "epoch": 0.3369095276220977, - "grad_norm": 0.36328125, - "learning_rate": 4.193636953425823e-05, - "loss": 1.6132, + "epoch": 0.40873218231708797, + "grad_norm": 0.80859375, + "learning_rate": 0.0001473527192433805, + "loss": 2.3611, "step": 1052 }, { - "epoch": 0.33722978382706165, - "grad_norm": 0.48046875, - "learning_rate": 4.191579251891572e-05, - "loss": 2.2119, + "epoch": 0.40912071100750347, + "grad_norm": 0.79296875, + "learning_rate": 0.0001472331487028378, + "loss": 2.369, "step": 1053 }, { - "epoch": 0.3375500400320256, - "grad_norm": 0.71484375, - "learning_rate": 4.1895194344893177e-05, - "loss": 2.0276, + "epoch": 0.40950923969791897, + "grad_norm": 0.7578125, + "learning_rate": 0.00014711349117725144, + "loss": 2.3826, "step": 1054 }, { - "epoch": 0.3378702962369896, - "grad_norm": 0.416015625, - "learning_rate": 4.187457503795527e-05, - "loss": 2.0666, + "epoch": 0.4098977683883344, + "grad_norm": 0.76171875, + "learning_rate": 0.0001469937468869839, + "loss": 2.4114, "step": 1055 }, { - "epoch": 0.33819055244195356, - "grad_norm": 0.490234375, - "learning_rate": 4.185393462389313e-05, - "loss": 2.1673, + "epoch": 0.4102862970787499, + "grad_norm": 0.7734375, + "learning_rate": 0.00014687391605255746, + "loss": 2.408, "step": 1056 }, { - "epoch": 0.3385108086469175, - "grad_norm": 0.369140625, - "learning_rate": 4.1833273128524284e-05, - "loss": 2.0761, + "epoch": 0.4106748257691654, + "grad_norm": 0.75390625, + "learning_rate": 0.0001467539988946538, + "loss": 2.3509, "step": 1057 }, { - "epoch": 0.3388310648518815, - "grad_norm": 0.40625, - "learning_rate": 4.181259057769264e-05, - "loss": 2.4819, + "epoch": 0.41106335445958087, + "grad_norm": 0.8125, + "learning_rate": 0.00014663399563411358, + "loss": 2.4604, "step": 1058 }, { - "epoch": 0.3391513210568455, - "grad_norm": 0.50390625, - "learning_rate": 4.1791886997268414e-05, - "loss": 2.1933, + "epoch": 0.41145188314999637, + "grad_norm": 0.75, + "learning_rate": 0.00014651390649193598, + "loss": 2.409, "step": 1059 }, { - "epoch": 0.33947157726180943, - "grad_norm": 0.419921875, - "learning_rate": 4.177116241314815e-05, - "loss": 2.2095, + "epoch": 0.4118404118404118, + "grad_norm": 0.7890625, + "learning_rate": 0.00014639373168927845, + "loss": 2.4373, "step": 1060 }, { - "epoch": 0.33979183346677344, - "grad_norm": 0.58984375, - "learning_rate": 4.175041685125465e-05, - "loss": 2.0342, + "epoch": 0.4122289405308273, + "grad_norm": 0.82421875, + "learning_rate": 0.0001462734714474561, + "loss": 2.4317, "step": 1061 }, { - "epoch": 0.3401120896717374, - "grad_norm": 0.55078125, - "learning_rate": 4.172965033753697e-05, - "loss": 2.3994, + "epoch": 0.4126174692212428, + "grad_norm": 0.78125, + "learning_rate": 0.00014615312598794135, + "loss": 2.4266, "step": 1062 }, { - "epoch": 0.34043234587670135, - "grad_norm": 0.310546875, - "learning_rate": 4.170886289797036e-05, - "loss": 2.0665, + "epoch": 0.41300599791165826, + "grad_norm": 0.74609375, + "learning_rate": 0.00014603269553236365, + "loss": 2.3566, "step": 1063 }, { - "epoch": 0.34075260208166536, - "grad_norm": 0.51953125, - "learning_rate": 4.1688054558556245e-05, - "loss": 1.9771, + "epoch": 0.41339452660207376, + "grad_norm": 0.84375, + "learning_rate": 0.00014591218030250892, + "loss": 2.4233, "step": 1064 }, { - "epoch": 0.3410728582866293, - "grad_norm": 0.5625, - "learning_rate": 4.16672253453222e-05, - "loss": 2.5785, + "epoch": 0.41378305529248927, + "grad_norm": 0.73828125, + "learning_rate": 0.0001457915805203193, + "loss": 2.4206, "step": 1065 }, { - "epoch": 0.34139311449159326, - "grad_norm": 0.53515625, - "learning_rate": 4.1646375284321916e-05, - "loss": 2.0384, + "epoch": 0.4141715839829047, + "grad_norm": 0.71875, + "learning_rate": 0.00014567089640789247, + "loss": 2.3427, "step": 1066 }, { - "epoch": 0.3417133706965573, - "grad_norm": 0.443359375, - "learning_rate": 4.1625504401635126e-05, - "loss": 2.1372, + "epoch": 0.4145601126733202, + "grad_norm": 0.7421875, + "learning_rate": 0.00014555012818748156, + "loss": 2.3342, "step": 1067 }, { - "epoch": 0.3420336269015212, - "grad_norm": 0.416015625, - "learning_rate": 4.1604612723367666e-05, - "loss": 1.9049, + "epoch": 0.4149486413637357, + "grad_norm": 0.734375, + "learning_rate": 0.00014542927608149456, + "loss": 2.3007, "step": 1068 }, { - "epoch": 0.3423538831064852, - "grad_norm": 0.33203125, - "learning_rate": 4.1583700275651314e-05, - "loss": 2.012, + "epoch": 0.41533717005415116, + "grad_norm": 0.78515625, + "learning_rate": 0.0001453083403124939, + "loss": 2.3456, "step": 1069 }, { - "epoch": 0.34267413931144913, - "grad_norm": 0.609375, - "learning_rate": 4.1562767084643885e-05, - "loss": 2.1901, + "epoch": 0.41572569874456666, + "grad_norm": 0.921875, + "learning_rate": 0.00014518732110319613, + "loss": 2.3655, "step": 1070 }, { - "epoch": 0.34299439551641314, - "grad_norm": 0.59765625, - "learning_rate": 4.154181317652911e-05, - "loss": 2.4813, + "epoch": 0.41611422743498216, + "grad_norm": 0.73828125, + "learning_rate": 0.00014506621867647142, + "loss": 2.3755, "step": 1071 }, { - "epoch": 0.3433146517213771, - "grad_norm": 0.5, - "learning_rate": 4.152083857751665e-05, - "loss": 2.4138, + "epoch": 0.4165027561253976, + "grad_norm": 0.74609375, + "learning_rate": 0.00014494503325534322, + "loss": 2.3675, "step": 1072 }, { - "epoch": 0.34363490792634105, - "grad_norm": 0.470703125, - "learning_rate": 4.149984331384203e-05, - "loss": 2.4764, + "epoch": 0.4168912848158131, + "grad_norm": 0.84375, + "learning_rate": 0.0001448237650629879, + "loss": 2.4362, "step": 1073 }, { - "epoch": 0.34395516413130506, - "grad_norm": 0.462890625, - "learning_rate": 4.147882741176662e-05, - "loss": 2.5071, + "epoch": 0.4172798135062286, + "grad_norm": 0.78515625, + "learning_rate": 0.0001447024143227341, + "loss": 2.3739, "step": 1074 }, { - "epoch": 0.344275420336269, - "grad_norm": 0.431640625, - "learning_rate": 4.145779089757764e-05, - "loss": 2.285, + "epoch": 0.41766834219664406, + "grad_norm": 0.8203125, + "learning_rate": 0.0001445809812580626, + "loss": 2.3791, "step": 1075 }, { - "epoch": 0.34459567654123296, - "grad_norm": 0.3984375, - "learning_rate": 4.1436733797588056e-05, - "loss": 2.2156, + "epoch": 0.41805687088705956, + "grad_norm": 0.74609375, + "learning_rate": 0.00014445946609260578, + "loss": 2.4234, "step": 1076 }, { - "epoch": 0.344915932746197, - "grad_norm": 3.4375, - "learning_rate": 4.141565613813658e-05, - "loss": 2.0952, + "epoch": 0.41844539957747506, + "grad_norm": 0.8671875, + "learning_rate": 0.00014433786905014716, + "loss": 2.4474, "step": 1077 }, { - "epoch": 0.3452361889511609, - "grad_norm": 0.515625, - "learning_rate": 4.139455794558768e-05, - "loss": 2.3122, + "epoch": 0.4188339282678905, + "grad_norm": 0.87109375, + "learning_rate": 0.00014421619035462116, + "loss": 2.4105, "step": 1078 }, { - "epoch": 0.3455564451561249, - "grad_norm": 0.361328125, - "learning_rate": 4.137343924633147e-05, - "loss": 1.9691, + "epoch": 0.419222456958306, + "grad_norm": 0.765625, + "learning_rate": 0.00014409443023011238, + "loss": 2.435, "step": 1079 }, { - "epoch": 0.3458767013610889, - "grad_norm": 0.345703125, - "learning_rate": 4.135230006678373e-05, - "loss": 2.2278, + "epoch": 0.4196109856487215, + "grad_norm": 0.89453125, + "learning_rate": 0.00014397258890085554, + "loss": 2.4015, "step": 1080 }, { - "epoch": 0.34619695756605284, - "grad_norm": 0.470703125, - "learning_rate": 4.133114043338585e-05, - "loss": 1.9843, + "epoch": 0.41999951433913696, + "grad_norm": 0.84765625, + "learning_rate": 0.00014385066659123487, + "loss": 2.3862, "step": 1081 }, { - "epoch": 0.3465172137710168, - "grad_norm": 0.5546875, - "learning_rate": 4.130996037260482e-05, - "loss": 1.5984, + "epoch": 0.42038804302955246, + "grad_norm": 0.84765625, + "learning_rate": 0.00014372866352578375, + "loss": 2.432, "step": 1082 }, { - "epoch": 0.3468374699759808, - "grad_norm": 0.302734375, - "learning_rate": 4.128875991093315e-05, - "loss": 2.2697, + "epoch": 0.42077657171996796, + "grad_norm": 0.765625, + "learning_rate": 0.00014360657992918422, + "loss": 2.3696, "step": 1083 }, { - "epoch": 0.34715772618094476, - "grad_norm": 0.5234375, - "learning_rate": 4.12675390748889e-05, - "loss": 2.314, + "epoch": 0.4211651004103834, + "grad_norm": 0.828125, + "learning_rate": 0.0001434844160262667, + "loss": 2.3393, "step": 1084 }, { - "epoch": 0.3474779823859087, - "grad_norm": 0.5078125, - "learning_rate": 4.12462978910156e-05, - "loss": 2.4867, + "epoch": 0.4215536291007989, + "grad_norm": 0.7734375, + "learning_rate": 0.00014336217204200942, + "loss": 2.4385, "step": 1085 }, { - "epoch": 0.3477982385908727, - "grad_norm": 0.48828125, - "learning_rate": 4.1225036385882235e-05, - "loss": 1.9026, + "epoch": 0.4219421577912144, + "grad_norm": 0.78515625, + "learning_rate": 0.0001432398482015382, + "loss": 2.3572, "step": 1086 }, { - "epoch": 0.3481184947958367, - "grad_norm": 0.435546875, - "learning_rate": 4.12037545860832e-05, - "loss": 2.6461, + "epoch": 0.42233068648162986, + "grad_norm": 0.828125, + "learning_rate": 0.0001431174447301258, + "loss": 2.4231, "step": 1087 }, { - "epoch": 0.34843875100080063, - "grad_norm": 0.6875, - "learning_rate": 4.118245251823829e-05, - "loss": 2.2835, + "epoch": 0.42271921517204536, + "grad_norm": 0.7578125, + "learning_rate": 0.0001429949618531917, + "loss": 2.3236, "step": 1088 }, { - "epoch": 0.34875900720576464, - "grad_norm": 0.392578125, - "learning_rate": 4.116113020899264e-05, - "loss": 2.3112, + "epoch": 0.42310774386246086, + "grad_norm": 0.8046875, + "learning_rate": 0.00014287239979630164, + "loss": 2.3515, "step": 1089 }, { - "epoch": 0.3490792634107286, - "grad_norm": 0.400390625, - "learning_rate": 4.1139787685016707e-05, - "loss": 2.0939, + "epoch": 0.4234962725528763, + "grad_norm": 0.7734375, + "learning_rate": 0.0001427497587851671, + "loss": 2.3919, "step": 1090 }, { - "epoch": 0.34939951961569254, - "grad_norm": 0.52734375, - "learning_rate": 4.1118424973006237e-05, - "loss": 1.9811, + "epoch": 0.4238848012432918, + "grad_norm": 0.8671875, + "learning_rate": 0.00014262703904564504, + "loss": 2.3854, "step": 1091 }, { - "epoch": 0.3497197758206565, - "grad_norm": 0.458984375, - "learning_rate": 4.1097042099682216e-05, - "loss": 2.5208, + "epoch": 0.4242733299337073, + "grad_norm": 0.79296875, + "learning_rate": 0.00014250424080373736, + "loss": 2.4189, "step": 1092 }, { - "epoch": 0.3500400320256205, - "grad_norm": 0.390625, - "learning_rate": 4.107563909179087e-05, - "loss": 1.6477, + "epoch": 0.42466185862412276, + "grad_norm": 0.8125, + "learning_rate": 0.0001423813642855905, + "loss": 2.4222, "step": 1093 }, { - "epoch": 0.35036028823058446, - "grad_norm": 0.55859375, - "learning_rate": 4.105421597610358e-05, - "loss": 2.1296, + "epoch": 0.42505038731453826, + "grad_norm": 0.76171875, + "learning_rate": 0.00014225840971749518, + "loss": 2.3939, "step": 1094 }, { - "epoch": 0.3506805444355484, - "grad_norm": 0.419921875, - "learning_rate": 4.1032772779416904e-05, - "loss": 2.5322, + "epoch": 0.42543891600495376, + "grad_norm": 0.75390625, + "learning_rate": 0.0001421353773258857, + "loss": 2.4524, "step": 1095 }, { - "epoch": 0.3510008006405124, - "grad_norm": 0.443359375, - "learning_rate": 4.10113095285525e-05, - "loss": 1.8264, + "epoch": 0.4258274446953692, + "grad_norm": 0.8125, + "learning_rate": 0.0001420122673373398, + "loss": 2.3831, "step": 1096 }, { - "epoch": 0.3513210568454764, - "grad_norm": 0.53515625, - "learning_rate": 4.098982625035713e-05, - "loss": 2.2612, + "epoch": 0.4262159733857847, + "grad_norm": 0.88671875, + "learning_rate": 0.00014188907997857804, + "loss": 2.2944, "step": 1097 }, { - "epoch": 0.35164131305044033, - "grad_norm": 0.439453125, - "learning_rate": 4.096832297170259e-05, - "loss": 1.9801, + "epoch": 0.4266045020762002, + "grad_norm": 0.7734375, + "learning_rate": 0.00014176581547646353, + "loss": 2.3749, "step": 1098 }, { - "epoch": 0.35196156925540434, - "grad_norm": 0.5546875, - "learning_rate": 4.094679971948569e-05, - "loss": 2.2899, + "epoch": 0.42699303076661566, + "grad_norm": 0.7578125, + "learning_rate": 0.00014164247405800144, + "loss": 2.3957, "step": 1099 }, { - "epoch": 0.3522818254603683, - "grad_norm": 0.50390625, - "learning_rate": 4.0925256520628244e-05, - "loss": 2.1428, + "epoch": 0.42738155945703116, + "grad_norm": 0.73828125, + "learning_rate": 0.00014151905595033852, + "loss": 2.3632, "step": 1100 }, { - "epoch": 0.35260208166533225, - "grad_norm": 0.40625, - "learning_rate": 4.090369340207699e-05, - "loss": 1.8895, + "epoch": 0.42777008814744666, + "grad_norm": 0.8984375, + "learning_rate": 0.00014139556138076286, + "loss": 2.384, "step": 1101 }, { - "epoch": 0.35292233787029625, - "grad_norm": 0.44140625, - "learning_rate": 4.088211039080361e-05, - "loss": 2.3364, + "epoch": 0.4281586168378621, + "grad_norm": 0.76953125, + "learning_rate": 0.00014127199057670326, + "loss": 2.4353, "step": 1102 }, { - "epoch": 0.3532425940752602, - "grad_norm": 0.5546875, - "learning_rate": 4.086050751380464e-05, - "loss": 1.9986, + "epoch": 0.4285471455282776, + "grad_norm": 0.75390625, + "learning_rate": 0.00014114834376572897, + "loss": 2.3899, "step": 1103 }, { - "epoch": 0.35356285028022416, - "grad_norm": 0.384765625, - "learning_rate": 4.083888479810149e-05, - "loss": 2.2253, + "epoch": 0.4289356742186931, + "grad_norm": 0.77734375, + "learning_rate": 0.00014102462117554924, + "loss": 2.4308, "step": 1104 }, { - "epoch": 0.35388310648518817, - "grad_norm": 0.384765625, - "learning_rate": 4.081724227074036e-05, - "loss": 2.1126, + "epoch": 0.42932420290910855, + "grad_norm": 0.7890625, + "learning_rate": 0.0001409008230340128, + "loss": 2.4151, "step": 1105 }, { - "epoch": 0.3542033626901521, - "grad_norm": 1.5078125, - "learning_rate": 4.079557995879225e-05, - "loss": 2.427, + "epoch": 0.42971273159952406, + "grad_norm": 0.734375, + "learning_rate": 0.0001407769495691076, + "loss": 2.3521, "step": 1106 }, { - "epoch": 0.3545236188951161, - "grad_norm": 0.5859375, - "learning_rate": 4.0773897889352913e-05, - "loss": 2.4259, + "epoch": 0.43010126028993956, + "grad_norm": 0.7734375, + "learning_rate": 0.00014065300100896022, + "loss": 2.3728, "step": 1107 }, { - "epoch": 0.3548438751000801, - "grad_norm": 0.3984375, - "learning_rate": 4.075219608954278e-05, - "loss": 2.0609, + "epoch": 0.430489788980355, + "grad_norm": 0.76953125, + "learning_rate": 0.0001405289775818356, + "loss": 2.4092, "step": 1108 }, { - "epoch": 0.35516413130504404, - "grad_norm": 0.86328125, - "learning_rate": 4.073047458650699e-05, - "loss": 2.5113, + "epoch": 0.4308783176707705, + "grad_norm": 0.76953125, + "learning_rate": 0.00014040487951613658, + "loss": 2.3365, "step": 1109 }, { - "epoch": 0.355484387510008, - "grad_norm": 0.6796875, - "learning_rate": 4.070873340741534e-05, - "loss": 2.1542, + "epoch": 0.431266846361186, + "grad_norm": 0.796875, + "learning_rate": 0.0001402807070404033, + "loss": 2.3233, "step": 1110 }, { - "epoch": 0.355804643714972, - "grad_norm": 0.373046875, - "learning_rate": 4.06869725794622e-05, - "loss": 2.0773, + "epoch": 0.43165537505160145, + "grad_norm": 0.7890625, + "learning_rate": 0.00014015646038331313, + "loss": 2.3724, "step": 1111 }, { - "epoch": 0.35612489991993596, - "grad_norm": 0.44140625, - "learning_rate": 4.066519212986654e-05, - "loss": 1.5918, + "epoch": 0.43204390374201695, + "grad_norm": 0.80859375, + "learning_rate": 0.00014003213977367994, + "loss": 2.4238, "step": 1112 }, { - "epoch": 0.3564451561248999, - "grad_norm": 0.337890625, - "learning_rate": 4.064339208587187e-05, - "loss": 1.8691, + "epoch": 0.43243243243243246, + "grad_norm": 0.8359375, + "learning_rate": 0.0001399077454404539, + "loss": 2.4004, "step": 1113 }, { - "epoch": 0.3567654123298639, - "grad_norm": 0.56640625, - "learning_rate": 4.0621572474746227e-05, - "loss": 2.0796, + "epoch": 0.4328209611228479, + "grad_norm": 0.78125, + "learning_rate": 0.00013978327761272072, + "loss": 2.4084, "step": 1114 }, { - "epoch": 0.35708566853482787, - "grad_norm": 0.439453125, - "learning_rate": 4.059973332378207e-05, - "loss": 2.437, + "epoch": 0.4332094898132634, + "grad_norm": 0.75390625, + "learning_rate": 0.00013965873651970175, + "loss": 2.3828, "step": 1115 }, { - "epoch": 0.3574059247397918, - "grad_norm": 0.494140625, - "learning_rate": 4.057787466029637e-05, - "loss": 2.6097, + "epoch": 0.4335980185036789, + "grad_norm": 0.703125, + "learning_rate": 0.00013953412239075302, + "loss": 2.3548, "step": 1116 }, { - "epoch": 0.3577261809447558, - "grad_norm": 0.5234375, - "learning_rate": 4.055599651163044e-05, - "loss": 2.567, + "epoch": 0.43398654719409435, + "grad_norm": 0.80859375, + "learning_rate": 0.00013940943545536524, + "loss": 2.4108, "step": 1117 }, { - "epoch": 0.3580464371497198, - "grad_norm": 0.47265625, - "learning_rate": 4.053409890515002e-05, - "loss": 2.0911, + "epoch": 0.43437507588450985, + "grad_norm": 0.7265625, + "learning_rate": 0.0001392846759431631, + "loss": 2.3684, "step": 1118 }, { - "epoch": 0.35836669335468374, - "grad_norm": 0.51953125, - "learning_rate": 4.051218186824515e-05, - "loss": 2.3121, + "epoch": 0.43476360457492536, + "grad_norm": 0.828125, + "learning_rate": 0.00013915984408390496, + "loss": 2.3744, "step": 1119 }, { - "epoch": 0.3586869495596477, - "grad_norm": 0.55859375, - "learning_rate": 4.049024542833018e-05, - "loss": 2.4189, + "epoch": 0.4351521332653408, + "grad_norm": 0.7890625, + "learning_rate": 0.00013903494010748246, + "loss": 2.3971, "step": 1120 }, { - "epoch": 0.3590072057646117, - "grad_norm": 0.5, - "learning_rate": 4.046828961284375e-05, - "loss": 2.6126, + "epoch": 0.4355406619557563, + "grad_norm": 0.79296875, + "learning_rate": 0.00013890996424392006, + "loss": 2.4105, "step": 1121 }, { - "epoch": 0.35932746196957566, - "grad_norm": 0.4609375, - "learning_rate": 4.0446314449248716e-05, - "loss": 2.208, + "epoch": 0.4359291906461718, + "grad_norm": 0.7265625, + "learning_rate": 0.0001387849167233745, + "loss": 2.2925, "step": 1122 }, { - "epoch": 0.3596477181745396, - "grad_norm": 0.53125, - "learning_rate": 4.042431996503213e-05, - "loss": 1.7206, + "epoch": 0.43631771933658725, + "grad_norm": 0.75, + "learning_rate": 0.00013865979777613458, + "loss": 2.3574, "step": 1123 }, { - "epoch": 0.3599679743795036, - "grad_norm": 0.427734375, - "learning_rate": 4.040230618770524e-05, - "loss": 2.3154, + "epoch": 0.43670624802700275, + "grad_norm": 0.78125, + "learning_rate": 0.00013853460763262062, + "loss": 2.3376, "step": 1124 }, { - "epoch": 0.3602882305844676, - "grad_norm": 0.53515625, - "learning_rate": 4.03802731448034e-05, - "loss": 2.3005, + "epoch": 0.43709477671741825, + "grad_norm": 0.76953125, + "learning_rate": 0.00013840934652338405, + "loss": 2.4241, "step": 1125 }, { - "epoch": 0.3606084867894315, - "grad_norm": 1.3046875, - "learning_rate": 4.0358220863886074e-05, - "loss": 2.2069, + "epoch": 0.4374833054078337, + "grad_norm": 0.7890625, + "learning_rate": 0.00013828401467910704, + "loss": 2.3839, "step": 1126 }, { - "epoch": 0.36092874299439553, - "grad_norm": 0.376953125, - "learning_rate": 4.033614937253678e-05, - "loss": 2.4196, + "epoch": 0.4378718340982492, + "grad_norm": 0.7734375, + "learning_rate": 0.00013815861233060193, + "loss": 2.3961, "step": 1127 }, { - "epoch": 0.3612489991993595, - "grad_norm": 0.427734375, - "learning_rate": 4.031405869836307e-05, - "loss": 2.4691, + "epoch": 0.43826036278866465, + "grad_norm": 0.73828125, + "learning_rate": 0.00013803313970881092, + "loss": 2.3902, "step": 1128 }, { - "epoch": 0.36156925540432344, - "grad_norm": 0.494140625, - "learning_rate": 4.0291948868996496e-05, - "loss": 2.7772, + "epoch": 0.43864889147908015, + "grad_norm": 0.85546875, + "learning_rate": 0.00013790759704480573, + "loss": 2.4001, "step": 1129 }, { - "epoch": 0.36188951160928745, - "grad_norm": 0.52734375, - "learning_rate": 4.026981991209256e-05, - "loss": 2.2456, + "epoch": 0.43903742016949565, + "grad_norm": 0.8046875, + "learning_rate": 0.00013778198456978696, + "loss": 2.3353, "step": 1130 }, { - "epoch": 0.3622097678142514, - "grad_norm": 0.5546875, - "learning_rate": 4.024767185533069e-05, - "loss": 2.1406, + "epoch": 0.4394259488599111, + "grad_norm": 0.78125, + "learning_rate": 0.00013765630251508386, + "loss": 2.3685, "step": 1131 }, { - "epoch": 0.36253002401921536, - "grad_norm": 0.5078125, - "learning_rate": 4.022550472641422e-05, - "loss": 2.3268, + "epoch": 0.4398144775503266, + "grad_norm": 0.73828125, + "learning_rate": 0.00013753055111215368, + "loss": 2.3767, "step": 1132 }, { - "epoch": 0.36285028022417937, - "grad_norm": 0.53125, - "learning_rate": 4.020331855307031e-05, - "loss": 2.2335, + "epoch": 0.4402030062407421, + "grad_norm": 0.8359375, + "learning_rate": 0.00013740473059258154, + "loss": 2.417, "step": 1133 }, { - "epoch": 0.3631705364291433, - "grad_norm": 0.43359375, - "learning_rate": 4.018111336304997e-05, - "loss": 2.384, + "epoch": 0.44059153493115755, + "grad_norm": 0.8359375, + "learning_rate": 0.00013727884118807976, + "loss": 2.4301, "step": 1134 }, { - "epoch": 0.3634907926341073, - "grad_norm": 0.4296875, - "learning_rate": 4.015888918412799e-05, - "loss": 2.0399, + "epoch": 0.44098006362157305, + "grad_norm": 0.75, + "learning_rate": 0.00013715288313048758, + "loss": 2.4261, "step": 1135 }, { - "epoch": 0.3638110488390713, - "grad_norm": 0.458984375, - "learning_rate": 4.01366460441029e-05, - "loss": 2.2721, + "epoch": 0.44136859231198855, + "grad_norm": 0.78125, + "learning_rate": 0.00013702685665177054, + "loss": 2.4326, "step": 1136 }, { - "epoch": 0.36413130504403524, - "grad_norm": 0.38671875, - "learning_rate": 4.011438397079696e-05, - "loss": 1.8505, + "epoch": 0.441757121002404, + "grad_norm": 0.91796875, + "learning_rate": 0.00013690076198402036, + "loss": 2.3514, "step": 1137 }, { - "epoch": 0.3644515612489992, - "grad_norm": 0.39453125, - "learning_rate": 4.0092102992056095e-05, - "loss": 1.8735, + "epoch": 0.4421456496928195, + "grad_norm": 0.94921875, + "learning_rate": 0.00013677459935945425, + "loss": 2.4182, "step": 1138 }, { - "epoch": 0.36477181745396314, - "grad_norm": 0.44140625, - "learning_rate": 4.006980313574991e-05, - "loss": 2.2612, + "epoch": 0.442534178383235, + "grad_norm": 0.7734375, + "learning_rate": 0.00013664836901041452, + "loss": 2.4883, "step": 1139 }, { - "epoch": 0.36509207365892715, - "grad_norm": 0.53515625, - "learning_rate": 4.004748442977158e-05, - "loss": 2.2203, + "epoch": 0.44292270707365045, + "grad_norm": 0.75390625, + "learning_rate": 0.00013652207116936828, + "loss": 2.3386, "step": 1140 }, { - "epoch": 0.3654123298638911, - "grad_norm": 0.609375, - "learning_rate": 4.002514690203788e-05, - "loss": 2.1066, + "epoch": 0.44331123576406595, + "grad_norm": 0.74609375, + "learning_rate": 0.00013639570606890694, + "loss": 2.3897, "step": 1141 }, { - "epoch": 0.36573258606885506, - "grad_norm": 0.462890625, - "learning_rate": 4.000279058048915e-05, - "loss": 2.2341, + "epoch": 0.44369976445448145, + "grad_norm": 0.8046875, + "learning_rate": 0.00013626927394174568, + "loss": 2.3805, "step": 1142 }, { - "epoch": 0.36605284227381907, - "grad_norm": 0.72265625, - "learning_rate": 3.998041549308919e-05, - "loss": 2.4673, + "epoch": 0.4440882931448969, + "grad_norm": 0.80078125, + "learning_rate": 0.00013614277502072326, + "loss": 2.358, "step": 1143 }, { - "epoch": 0.366373098478783, - "grad_norm": 0.427734375, - "learning_rate": 3.995802166782531e-05, - "loss": 2.1514, + "epoch": 0.4444768218353124, + "grad_norm": 0.84765625, + "learning_rate": 0.0001360162095388013, + "loss": 2.3745, "step": 1144 }, { - "epoch": 0.366693354683747, - "grad_norm": 0.54296875, - "learning_rate": 3.993560913270823e-05, - "loss": 1.8712, + "epoch": 0.4448653505257279, + "grad_norm": 0.72265625, + "learning_rate": 0.00013588957772906412, + "loss": 2.3668, "step": 1145 }, { - "epoch": 0.367013610888711, - "grad_norm": 0.46484375, - "learning_rate": 3.991317791577212e-05, - "loss": 2.6296, + "epoch": 0.44525387921614334, + "grad_norm": 0.8046875, + "learning_rate": 0.00013576287982471812, + "loss": 2.3509, "step": 1146 }, { - "epoch": 0.36733386709367494, - "grad_norm": 0.48046875, - "learning_rate": 3.989072804507444e-05, - "loss": 2.4734, + "epoch": 0.44564240790655885, + "grad_norm": 0.74609375, + "learning_rate": 0.00013563611605909141, + "loss": 2.3841, "step": 1147 }, { - "epoch": 0.3676541232986389, - "grad_norm": 0.5546875, - "learning_rate": 3.986825954869607e-05, - "loss": 2.5021, + "epoch": 0.44603093659697435, + "grad_norm": 0.8125, + "learning_rate": 0.00013550928666563347, + "loss": 2.3929, "step": 1148 }, { - "epoch": 0.3679743795036029, - "grad_norm": 0.48046875, - "learning_rate": 3.9845772454741124e-05, - "loss": 2.0079, + "epoch": 0.4464194652873898, + "grad_norm": 0.734375, + "learning_rate": 0.0001353823918779146, + "loss": 2.377, "step": 1149 }, { - "epoch": 0.36829463570856685, - "grad_norm": 0.65625, - "learning_rate": 3.982326679133701e-05, - "loss": 2.6182, + "epoch": 0.4468079939778053, + "grad_norm": 0.7265625, + "learning_rate": 0.00013525543192962548, + "loss": 2.4199, "step": 1150 }, { - "epoch": 0.3686148919135308, - "grad_norm": 0.43359375, - "learning_rate": 3.9800742586634346e-05, - "loss": 1.8996, + "epoch": 0.4471965226682208, + "grad_norm": 0.7734375, + "learning_rate": 0.00013512840705457685, + "loss": 2.4202, "step": 1151 }, { - "epoch": 0.3689351481184948, - "grad_norm": 0.65625, - "learning_rate": 3.977819986880694e-05, - "loss": 2.4005, + "epoch": 0.44758505135863624, + "grad_norm": 0.7421875, + "learning_rate": 0.000135001317486699, + "loss": 2.3966, "step": 1152 }, { - "epoch": 0.36925540432345877, - "grad_norm": 0.95703125, - "learning_rate": 3.9755638666051795e-05, - "loss": 2.2208, + "epoch": 0.44797358004905175, + "grad_norm": 0.70703125, + "learning_rate": 0.00013487416346004137, + "loss": 2.4289, "step": 1153 }, { - "epoch": 0.3695756605284227, - "grad_norm": 1.2421875, - "learning_rate": 3.973305900658897e-05, - "loss": 1.9195, + "epoch": 0.44836210873946725, + "grad_norm": 0.75, + "learning_rate": 0.00013474694520877208, + "loss": 2.3513, "step": 1154 }, { - "epoch": 0.36989591673338673, - "grad_norm": 0.462890625, - "learning_rate": 3.971046091866166e-05, - "loss": 1.4788, + "epoch": 0.4487506374298827, + "grad_norm": 0.79296875, + "learning_rate": 0.0001346196629671776, + "loss": 2.3925, "step": 1155 }, { - "epoch": 0.3702161729383507, - "grad_norm": 0.455078125, - "learning_rate": 3.968784443053612e-05, - "loss": 2.1755, + "epoch": 0.4491391661202982, + "grad_norm": 0.88671875, + "learning_rate": 0.00013449231696966208, + "loss": 2.4184, "step": 1156 }, { - "epoch": 0.37053642914331464, - "grad_norm": 0.53515625, - "learning_rate": 3.966520957050156e-05, - "loss": 2.3551, + "epoch": 0.4495276948107137, + "grad_norm": 0.75, + "learning_rate": 0.00013436490745074735, + "loss": 2.364, "step": 1157 }, { - "epoch": 0.37085668534827865, - "grad_norm": 0.53125, - "learning_rate": 3.964255636687023e-05, - "loss": 2.3383, + "epoch": 0.44991622350112914, + "grad_norm": 0.734375, + "learning_rate": 0.00013423743464507194, + "loss": 2.3693, "step": 1158 }, { - "epoch": 0.3711769415532426, - "grad_norm": 0.451171875, - "learning_rate": 3.961988484797729e-05, - "loss": 1.8462, + "epoch": 0.45030475219154464, + "grad_norm": 0.84765625, + "learning_rate": 0.0001341098987873911, + "loss": 2.3543, "step": 1159 }, { - "epoch": 0.37149719775820655, - "grad_norm": 0.52734375, - "learning_rate": 3.959719504218083e-05, - "loss": 2.2438, + "epoch": 0.45069328088196015, + "grad_norm": 0.7578125, + "learning_rate": 0.00013398230011257614, + "loss": 2.426, "step": 1160 }, { - "epoch": 0.37181745396317056, - "grad_norm": 0.5078125, - "learning_rate": 3.95744869778618e-05, - "loss": 2.2278, + "epoch": 0.4510818095723756, + "grad_norm": 0.77734375, + "learning_rate": 0.00013385463885561412, + "loss": 2.3451, "step": 1161 }, { - "epoch": 0.3721377101681345, - "grad_norm": 0.455078125, - "learning_rate": 3.9551760683424e-05, - "loss": 2.2491, + "epoch": 0.4514703382627911, + "grad_norm": 0.7265625, + "learning_rate": 0.00013372691525160725, + "loss": 2.3361, "step": 1162 }, { - "epoch": 0.37245796637309847, - "grad_norm": 0.80078125, - "learning_rate": 3.952901618729402e-05, - "loss": 1.8948, + "epoch": 0.4518588669532066, + "grad_norm": 0.89453125, + "learning_rate": 0.0001335991295357726, + "loss": 2.4234, "step": 1163 }, { - "epoch": 0.3727782225780624, - "grad_norm": 0.55078125, - "learning_rate": 3.950625351792122e-05, - "loss": 1.8903, + "epoch": 0.45224739564362204, + "grad_norm": 0.78125, + "learning_rate": 0.00013347128194344168, + "loss": 2.3376, "step": 1164 }, { - "epoch": 0.37309847878302643, - "grad_norm": 0.359375, - "learning_rate": 3.948347270377769e-05, - "loss": 1.2149, + "epoch": 0.45263592433403754, + "grad_norm": 0.7265625, + "learning_rate": 0.0001333433727100599, + "loss": 2.3383, "step": 1165 }, { - "epoch": 0.3734187349879904, - "grad_norm": 0.38671875, - "learning_rate": 3.946067377335824e-05, - "loss": 2.2527, + "epoch": 0.45302445302445304, + "grad_norm": 1.3203125, + "learning_rate": 0.00013321540207118614, + "loss": 2.4261, "step": 1166 }, { - "epoch": 0.37373899119295434, - "grad_norm": 0.62890625, - "learning_rate": 3.94378567551803e-05, - "loss": 1.5419, + "epoch": 0.4534129817148685, + "grad_norm": 0.81640625, + "learning_rate": 0.00013308737026249248, + "loss": 2.4405, "step": 1167 }, { - "epoch": 0.37405924739791835, - "grad_norm": 0.453125, - "learning_rate": 3.941502167778395e-05, - "loss": 1.9236, + "epoch": 0.453801510405284, + "grad_norm": 0.76953125, + "learning_rate": 0.00013295927751976358, + "loss": 2.356, "step": 1168 }, { - "epoch": 0.3743795036028823, - "grad_norm": 0.443359375, - "learning_rate": 3.9392168569731854e-05, - "loss": 2.3081, + "epoch": 0.4541900390956995, + "grad_norm": 0.8515625, + "learning_rate": 0.00013283112407889633, + "loss": 2.4666, "step": 1169 }, { - "epoch": 0.37469975980784626, - "grad_norm": 0.515625, - "learning_rate": 3.9369297459609247e-05, - "loss": 2.1903, + "epoch": 0.45457856778611494, + "grad_norm": 0.76953125, + "learning_rate": 0.00013270291017589936, + "loss": 2.4644, "step": 1170 }, { - "epoch": 0.37502001601281026, - "grad_norm": 0.859375, - "learning_rate": 3.934640837602383e-05, - "loss": 2.5142, + "epoch": 0.45496709647653044, + "grad_norm": 0.796875, + "learning_rate": 0.00013257463604689275, + "loss": 2.3546, "step": 1171 }, { - "epoch": 0.3753402722177742, - "grad_norm": 0.58203125, - "learning_rate": 3.932350134760585e-05, - "loss": 2.1751, + "epoch": 0.45535562516694594, + "grad_norm": 0.86328125, + "learning_rate": 0.0001324463019281074, + "loss": 2.4329, "step": 1172 }, { - "epoch": 0.37566052842273817, - "grad_norm": 0.6484375, - "learning_rate": 3.930057640300795e-05, - "loss": 2.3225, + "epoch": 0.4557441538573614, + "grad_norm": 0.75390625, + "learning_rate": 0.00013231790805588468, + "loss": 2.3792, "step": 1173 }, { - "epoch": 0.3759807846277022, - "grad_norm": 0.5078125, - "learning_rate": 3.92776335709052e-05, - "loss": 2.1874, + "epoch": 0.4561326825477769, + "grad_norm": 0.7890625, + "learning_rate": 0.0001321894546666761, + "loss": 2.352, "step": 1174 }, { - "epoch": 0.37630104083266613, - "grad_norm": 0.51953125, - "learning_rate": 3.925467287999507e-05, - "loss": 2.1818, + "epoch": 0.4565212112381924, + "grad_norm": 0.76953125, + "learning_rate": 0.00013206094199704274, + "loss": 2.3812, "step": 1175 }, { - "epoch": 0.3766212970376301, - "grad_norm": 0.5234375, - "learning_rate": 3.9231694358997326e-05, - "loss": 2.558, + "epoch": 0.45690973992860784, + "grad_norm": 0.72265625, + "learning_rate": 0.00013193237028365478, + "loss": 2.3032, "step": 1176 }, { - "epoch": 0.3769415532425941, - "grad_norm": 0.38671875, - "learning_rate": 3.920869803665405e-05, - "loss": 2.0283, + "epoch": 0.45729826861902334, + "grad_norm": 0.6953125, + "learning_rate": 0.00013180373976329118, + "loss": 2.3755, "step": 1177 }, { - "epoch": 0.37726180944755805, - "grad_norm": 0.5, - "learning_rate": 3.91856839417296e-05, - "loss": 2.6307, + "epoch": 0.45768679730943884, + "grad_norm": 0.69921875, + "learning_rate": 0.00013167505067283926, + "loss": 2.4432, "step": 1178 }, { - "epoch": 0.377582065652522, - "grad_norm": 0.578125, - "learning_rate": 3.916265210301057e-05, - "loss": 1.9584, + "epoch": 0.4580753259998543, + "grad_norm": 0.66015625, + "learning_rate": 0.0001315463032492941, + "loss": 2.2993, "step": 1179 }, { - "epoch": 0.377902321857486, - "grad_norm": 0.5, - "learning_rate": 3.913960254930572e-05, - "loss": 1.7026, + "epoch": 0.4584638546902698, + "grad_norm": 0.71484375, + "learning_rate": 0.00013141749772975825, + "loss": 2.3378, "step": 1180 }, { - "epoch": 0.37822257806244997, - "grad_norm": 0.50390625, - "learning_rate": 3.9116535309446e-05, - "loss": 2.0822, + "epoch": 0.4588523833806853, + "grad_norm": 14.75, + "learning_rate": 0.00013128863435144127, + "loss": 2.37, "step": 1181 }, { - "epoch": 0.3785428342674139, - "grad_norm": 0.60546875, - "learning_rate": 3.909345041228447e-05, - "loss": 2.5519, + "epoch": 0.45924091207110074, + "grad_norm": 0.79296875, + "learning_rate": 0.00013115971335165926, + "loss": 2.3849, "step": 1182 }, { - "epoch": 0.37886309047237793, - "grad_norm": 0.447265625, - "learning_rate": 3.9070347886696254e-05, - "loss": 2.2184, + "epoch": 0.45962944076151624, + "grad_norm": 0.8046875, + "learning_rate": 0.00013103073496783447, + "loss": 2.4525, "step": 1183 }, { - "epoch": 0.3791833466773419, - "grad_norm": 0.921875, - "learning_rate": 3.9047227761578566e-05, - "loss": 1.765, + "epoch": 0.46001796945193174, + "grad_norm": 0.78515625, + "learning_rate": 0.00013090169943749476, + "loss": 2.3965, "step": 1184 }, { - "epoch": 0.37950360288230583, - "grad_norm": 0.65625, - "learning_rate": 3.902409006585061e-05, - "loss": 2.3076, + "epoch": 0.4604064981423472, + "grad_norm": 0.74609375, + "learning_rate": 0.00013077260699827326, + "loss": 2.4148, "step": 1185 }, { - "epoch": 0.3798238590872698, - "grad_norm": 0.490234375, - "learning_rate": 3.900093482845357e-05, - "loss": 2.0427, + "epoch": 0.4607950268327627, + "grad_norm": 0.68359375, + "learning_rate": 0.00013064345788790788, + "loss": 2.3294, "step": 1186 }, { - "epoch": 0.3801441152922338, - "grad_norm": 1.640625, - "learning_rate": 3.8977762078350576e-05, - "loss": 1.6712, + "epoch": 0.4611835555231782, + "grad_norm": 0.70703125, + "learning_rate": 0.00013051425234424105, + "loss": 2.4082, "step": 1187 }, { - "epoch": 0.38046437149719775, - "grad_norm": 3.296875, - "learning_rate": 3.895457184452665e-05, - "loss": 2.4834, + "epoch": 0.46157208421359364, + "grad_norm": 0.74609375, + "learning_rate": 0.00013038499060521886, + "loss": 2.4042, "step": 1188 }, { - "epoch": 0.3807846277021617, - "grad_norm": 0.349609375, - "learning_rate": 3.8931364155988726e-05, - "loss": 2.2172, + "epoch": 0.46196061290400914, + "grad_norm": 0.83203125, + "learning_rate": 0.00013025567290889112, + "loss": 2.4304, "step": 1189 }, { - "epoch": 0.3811048839071257, - "grad_norm": 0.58984375, - "learning_rate": 3.89081390417655e-05, - "loss": 2.2178, + "epoch": 0.46234914159442464, + "grad_norm": 0.8046875, + "learning_rate": 0.00013012629949341053, + "loss": 2.3919, "step": 1190 }, { - "epoch": 0.38142514011208967, - "grad_norm": 0.6171875, - "learning_rate": 3.888489653090752e-05, - "loss": 2.2411, + "epoch": 0.4627376702848401, + "grad_norm": 0.7109375, + "learning_rate": 0.00012999687059703257, + "loss": 2.3465, "step": 1191 }, { - "epoch": 0.3817453963170536, - "grad_norm": 0.400390625, - "learning_rate": 3.886163665248709e-05, - "loss": 2.7614, + "epoch": 0.4631261989752556, + "grad_norm": 0.75, + "learning_rate": 0.0001298673864581147, + "loss": 2.3778, "step": 1192 }, { - "epoch": 0.38206565252201763, - "grad_norm": 0.56640625, - "learning_rate": 3.8838359435598205e-05, - "loss": 2.2614, + "epoch": 0.4635147276656711, + "grad_norm": 0.7890625, + "learning_rate": 0.00012973784731511637, + "loss": 2.4215, "step": 1193 }, { - "epoch": 0.3823859087269816, - "grad_norm": 0.51171875, - "learning_rate": 3.881506490935657e-05, - "loss": 2.0497, + "epoch": 0.46390325635608654, + "grad_norm": 0.7734375, + "learning_rate": 0.00012960825340659803, + "loss": 2.3797, "step": 1194 }, { - "epoch": 0.38270616493194554, - "grad_norm": 0.61328125, - "learning_rate": 3.879175310289955e-05, - "loss": 2.3673, + "epoch": 0.46429178504650204, + "grad_norm": 0.78125, + "learning_rate": 0.00012947860497122119, + "loss": 2.4168, "step": 1195 }, { - "epoch": 0.38302642113690955, - "grad_norm": 4.6875, - "learning_rate": 3.876842404538611e-05, - "loss": 2.7132, + "epoch": 0.4646803137369175, + "grad_norm": 0.7421875, + "learning_rate": 0.0001293489022477477, + "loss": 2.3733, "step": 1196 }, { - "epoch": 0.3833466773418735, - "grad_norm": 0.498046875, - "learning_rate": 3.874507776599678e-05, - "loss": 2.1813, + "epoch": 0.465068842427333, + "grad_norm": 0.78515625, + "learning_rate": 0.00012921914547503943, + "loss": 2.3437, "step": 1197 }, { - "epoch": 0.38366693354683745, - "grad_norm": 1.3359375, - "learning_rate": 3.872171429393368e-05, - "loss": 2.3327, + "epoch": 0.4654573711177485, + "grad_norm": 0.75390625, + "learning_rate": 0.00012908933489205775, + "loss": 2.3236, "step": 1198 }, { - "epoch": 0.38398718975180146, - "grad_norm": 0.640625, - "learning_rate": 3.8698333658420366e-05, - "loss": 1.5882, + "epoch": 0.46584589980816393, + "grad_norm": 0.76171875, + "learning_rate": 0.00012895947073786313, + "loss": 2.4033, "step": 1199 }, { - "epoch": 0.3843074459567654, - "grad_norm": 0.53515625, - "learning_rate": 3.867493588870193e-05, - "loss": 2.0014, + "epoch": 0.46623442849857943, + "grad_norm": 0.74609375, + "learning_rate": 0.00012882955325161472, + "loss": 2.3573, "step": 1200 }, { - "epoch": 0.38462770216172937, - "grad_norm": 0.494140625, - "learning_rate": 3.865152101404485e-05, - "loss": 2.2738, + "epoch": 0.46662295718899494, + "grad_norm": 0.703125, + "learning_rate": 0.00012869958267256988, + "loss": 2.3784, "step": 1201 }, { - "epoch": 0.3849479583666934, - "grad_norm": 0.44921875, - "learning_rate": 3.862808906373701e-05, - "loss": 2.3236, + "epoch": 0.4670114858794104, + "grad_norm": 0.734375, + "learning_rate": 0.00012856955924008375, + "loss": 2.3903, "step": 1202 }, { - "epoch": 0.38526821457165733, - "grad_norm": 0.43359375, - "learning_rate": 3.860464006708766e-05, - "loss": 2.6712, + "epoch": 0.4674000145698259, + "grad_norm": 0.7265625, + "learning_rate": 0.00012843948319360874, + "loss": 2.3779, "step": 1203 }, { - "epoch": 0.3855884707766213, - "grad_norm": 0.470703125, - "learning_rate": 3.8581174053427374e-05, - "loss": 1.9265, + "epoch": 0.4677885432602414, + "grad_norm": 0.7890625, + "learning_rate": 0.00012830935477269425, + "loss": 2.4628, "step": 1204 }, { - "epoch": 0.3859087269815853, - "grad_norm": 0.46484375, - "learning_rate": 3.855769105210801e-05, - "loss": 2.5721, + "epoch": 0.46817707195065683, + "grad_norm": 0.765625, + "learning_rate": 0.00012817917421698613, + "loss": 2.3718, "step": 1205 }, { - "epoch": 0.38622898318654925, - "grad_norm": 0.462890625, - "learning_rate": 3.8534191092502655e-05, - "loss": 2.0682, + "epoch": 0.46856560064107233, + "grad_norm": 0.73828125, + "learning_rate": 0.00012804894176622617, + "loss": 2.368, "step": 1206 }, { - "epoch": 0.3865492393915132, - "grad_norm": 0.671875, - "learning_rate": 3.851067420400564e-05, - "loss": 2.3802, + "epoch": 0.46895412933148783, + "grad_norm": 0.72265625, + "learning_rate": 0.00012791865766025176, + "loss": 2.4149, "step": 1207 }, { - "epoch": 0.3868694955964772, - "grad_norm": 0.5234375, - "learning_rate": 3.848714041603245e-05, - "loss": 2.3584, + "epoch": 0.4693426580219033, + "grad_norm": 0.76171875, + "learning_rate": 0.0001277883221389954, + "loss": 2.3687, "step": 1208 }, { - "epoch": 0.38718975180144116, - "grad_norm": 0.6328125, - "learning_rate": 3.846358975801971e-05, - "loss": 2.3271, + "epoch": 0.4697311867123188, + "grad_norm": 0.7734375, + "learning_rate": 0.00012765793544248442, + "loss": 2.4661, "step": 1209 }, { - "epoch": 0.3875100080064051, - "grad_norm": 0.60546875, - "learning_rate": 3.844002225942514e-05, - "loss": 2.2613, + "epoch": 0.4701197154027343, + "grad_norm": 0.73046875, + "learning_rate": 0.0001275274978108401, + "loss": 2.3438, "step": 1210 }, { - "epoch": 0.38783026421136907, - "grad_norm": 0.435546875, - "learning_rate": 3.8416437949727555e-05, - "loss": 2.1979, + "epoch": 0.47050824409314973, + "grad_norm": 0.73046875, + "learning_rate": 0.00012739700948427785, + "loss": 2.4013, "step": 1211 }, { - "epoch": 0.3881505204163331, - "grad_norm": 0.578125, - "learning_rate": 3.839283685842676e-05, - "loss": 2.7155, + "epoch": 0.47089677278356523, + "grad_norm": 0.73046875, + "learning_rate": 0.00012726647070310622, + "loss": 2.4636, "step": 1212 }, { - "epoch": 0.38847077662129703, - "grad_norm": 0.4296875, - "learning_rate": 3.836921901504358e-05, - "loss": 1.9949, + "epoch": 0.47128530147398073, + "grad_norm": 0.7734375, + "learning_rate": 0.00012713588170772675, + "loss": 2.4221, "step": 1213 }, { - "epoch": 0.388791032826261, - "grad_norm": 0.53125, - "learning_rate": 3.8345584449119776e-05, - "loss": 2.1843, + "epoch": 0.4716738301643962, + "grad_norm": 0.70703125, + "learning_rate": 0.00012700524273863347, + "loss": 2.4232, "step": 1214 }, { - "epoch": 0.389111289031225, - "grad_norm": 0.5390625, - "learning_rate": 3.832193319021802e-05, - "loss": 2.2712, + "epoch": 0.4720623588548117, + "grad_norm": 0.8046875, + "learning_rate": 0.00012687455403641242, + "loss": 2.3794, "step": 1215 }, { - "epoch": 0.38943154523618895, - "grad_norm": 0.609375, - "learning_rate": 3.8298265267921884e-05, - "loss": 2.1854, + "epoch": 0.4724508875452272, + "grad_norm": 0.74609375, + "learning_rate": 0.00012674381584174124, + "loss": 2.4058, "step": 1216 }, { - "epoch": 0.3897518014411529, - "grad_norm": 1.0625, - "learning_rate": 3.827458071183578e-05, - "loss": 2.5162, + "epoch": 0.47283941623564263, + "grad_norm": 0.78515625, + "learning_rate": 0.00012661302839538866, + "loss": 2.3517, "step": 1217 }, { - "epoch": 0.3900720576461169, - "grad_norm": 0.369140625, - "learning_rate": 3.825087955158492e-05, - "loss": 2.4592, + "epoch": 0.47322794492605813, + "grad_norm": 0.75, + "learning_rate": 0.00012648219193821424, + "loss": 2.4763, "step": 1218 }, { - "epoch": 0.39039231385108086, - "grad_norm": 0.71875, - "learning_rate": 3.822716181681528e-05, - "loss": 1.9011, + "epoch": 0.47361647361647363, + "grad_norm": 0.75390625, + "learning_rate": 0.00012635130671116772, + "loss": 2.3852, "step": 1219 }, { - "epoch": 0.3907125700560448, - "grad_norm": 0.416015625, - "learning_rate": 3.820342753719357e-05, - "loss": 2.2409, + "epoch": 0.4740050023068891, + "grad_norm": 0.87890625, + "learning_rate": 0.00012622037295528858, + "loss": 2.3484, "step": 1220 }, { - "epoch": 0.3910328262610088, - "grad_norm": 0.4375, - "learning_rate": 3.817967674240721e-05, - "loss": 1.9098, + "epoch": 0.4743935309973046, + "grad_norm": 0.72265625, + "learning_rate": 0.00012608939091170577, + "loss": 2.3919, "step": 1221 }, { - "epoch": 0.3913530824659728, - "grad_norm": 0.484375, - "learning_rate": 3.815590946216425e-05, - "loss": 2.0723, + "epoch": 0.4747820596877201, + "grad_norm": 0.71484375, + "learning_rate": 0.00012595836082163718, + "loss": 2.4426, "step": 1222 }, { - "epoch": 0.39167333867093673, - "grad_norm": 0.373046875, - "learning_rate": 3.813212572619338e-05, - "loss": 2.3035, + "epoch": 0.4751705883781355, + "grad_norm": 0.69921875, + "learning_rate": 0.00012582728292638912, + "loss": 2.3485, "step": 1223 }, { - "epoch": 0.39199359487590074, - "grad_norm": 0.439453125, - "learning_rate": 3.810832556424388e-05, - "loss": 1.9079, + "epoch": 0.47555911706855103, + "grad_norm": 0.72265625, + "learning_rate": 0.000125696157467356, + "loss": 2.4109, "step": 1224 }, { - "epoch": 0.3923138510808647, - "grad_norm": 1.546875, - "learning_rate": 3.808450900608555e-05, - "loss": 1.7984, + "epoch": 0.47594764575896653, + "grad_norm": 0.76953125, + "learning_rate": 0.00012556498468601975, + "loss": 2.4374, "step": 1225 }, { - "epoch": 0.39263410728582865, - "grad_norm": 0.486328125, - "learning_rate": 3.806067608150872e-05, - "loss": 2.1234, + "epoch": 0.476336174449382, + "grad_norm": 0.76171875, + "learning_rate": 0.0001254337648239495, + "loss": 2.3384, "step": 1226 }, { - "epoch": 0.39295436349079266, - "grad_norm": 0.5, - "learning_rate": 3.803682682032419e-05, - "loss": 2.2432, + "epoch": 0.4767247031397975, + "grad_norm": 0.734375, + "learning_rate": 0.00012530249812280108, + "loss": 2.3056, "step": 1227 }, { - "epoch": 0.3932746196957566, - "grad_norm": 1.0703125, - "learning_rate": 3.8012961252363184e-05, - "loss": 1.9167, + "epoch": 0.477113231830213, + "grad_norm": 0.765625, + "learning_rate": 0.0001251711848243166, + "loss": 2.416, "step": 1228 }, { - "epoch": 0.39359487590072056, - "grad_norm": 0.42578125, - "learning_rate": 3.798907940747732e-05, - "loss": 2.6162, + "epoch": 0.4775017605206284, + "grad_norm": 0.73046875, + "learning_rate": 0.00012503982517032388, + "loss": 2.3606, "step": 1229 }, { - "epoch": 0.3939151321056846, - "grad_norm": 0.45703125, - "learning_rate": 3.79651813155386e-05, - "loss": 2.239, + "epoch": 0.47789028921104393, + "grad_norm": 0.734375, + "learning_rate": 0.00012490841940273627, + "loss": 2.4288, "step": 1230 }, { - "epoch": 0.3942353883106485, - "grad_norm": 0.65625, - "learning_rate": 3.794126700643933e-05, - "loss": 1.7196, + "epoch": 0.47827881790145943, + "grad_norm": 0.76171875, + "learning_rate": 0.0001247769677635519, + "loss": 2.4206, "step": 1231 }, { - "epoch": 0.3945556445156125, - "grad_norm": 0.408203125, - "learning_rate": 3.791733651009209e-05, - "loss": 2.1263, + "epoch": 0.4786673465918749, + "grad_norm": 0.71484375, + "learning_rate": 0.00012464547049485347, + "loss": 2.3527, "step": 1232 }, { - "epoch": 0.39487590072057643, - "grad_norm": 0.427734375, - "learning_rate": 3.789338985642973e-05, - "loss": 2.0495, + "epoch": 0.4790558752822904, + "grad_norm": 0.75390625, + "learning_rate": 0.00012451392783880766, + "loss": 2.4884, "step": 1233 }, { - "epoch": 0.39519615692554044, - "grad_norm": 0.45703125, - "learning_rate": 3.786942707540529e-05, - "loss": 2.2755, + "epoch": 0.4794444039727059, + "grad_norm": 0.765625, + "learning_rate": 0.00012438234003766478, + "loss": 2.4246, "step": 1234 }, { - "epoch": 0.3955164131305044, - "grad_norm": 0.494140625, - "learning_rate": 3.784544819699202e-05, - "loss": 2.5356, + "epoch": 0.4798329326631213, + "grad_norm": 0.7578125, + "learning_rate": 0.0001242507073337582, + "loss": 2.4027, "step": 1235 }, { - "epoch": 0.39583666933546835, - "grad_norm": 0.39453125, - "learning_rate": 3.7821453251183245e-05, - "loss": 2.3301, + "epoch": 0.4802214613535368, + "grad_norm": 0.73828125, + "learning_rate": 0.00012411902996950407, + "loss": 2.3954, "step": 1236 }, { - "epoch": 0.39615692554043236, - "grad_norm": 0.48046875, - "learning_rate": 3.7797442267992436e-05, - "loss": 2.3236, + "epoch": 0.48060999004395233, + "grad_norm": 0.76171875, + "learning_rate": 0.00012398730818740077, + "loss": 2.4057, "step": 1237 }, { - "epoch": 0.3964771817453963, - "grad_norm": 0.5625, - "learning_rate": 3.7773415277453103e-05, - "loss": 2.3126, + "epoch": 0.4809985187343678, + "grad_norm": 0.75, + "learning_rate": 0.00012385554223002845, + "loss": 2.3871, "step": 1238 }, { - "epoch": 0.39679743795036027, - "grad_norm": 0.349609375, - "learning_rate": 3.77493723096188e-05, - "loss": 2.2927, + "epoch": 0.4813870474247833, + "grad_norm": 0.71484375, + "learning_rate": 0.0001237237323400486, + "loss": 2.4078, "step": 1239 }, { - "epoch": 0.3971176941553243, - "grad_norm": 0.5078125, - "learning_rate": 3.772531339456302e-05, - "loss": 1.9049, + "epoch": 0.4817755761151988, + "grad_norm": 0.75, + "learning_rate": 0.00012359187876020367, + "loss": 2.4764, "step": 1240 }, { - "epoch": 0.39743795036028823, - "grad_norm": 0.6796875, - "learning_rate": 3.770123856237925e-05, - "loss": 2.5726, + "epoch": 0.4821641048056142, + "grad_norm": 0.703125, + "learning_rate": 0.0001234599817333166, + "loss": 2.3753, "step": 1241 }, { - "epoch": 0.3977582065652522, - "grad_norm": 0.490234375, - "learning_rate": 3.767714784318088e-05, - "loss": 2.2026, + "epoch": 0.4825526334960297, + "grad_norm": 0.73828125, + "learning_rate": 0.00012332804150229018, + "loss": 2.3435, "step": 1242 }, { - "epoch": 0.3980784627702162, - "grad_norm": 0.62890625, - "learning_rate": 3.765304126710113e-05, - "loss": 2.08, + "epoch": 0.4829411621864452, + "grad_norm": 0.703125, + "learning_rate": 0.00012319605831010694, + "loss": 2.3939, "step": 1243 }, { - "epoch": 0.39839871897518014, - "grad_norm": 0.453125, - "learning_rate": 3.762891886429312e-05, - "loss": 1.8371, + "epoch": 0.4833296908768607, + "grad_norm": 0.78515625, + "learning_rate": 0.00012306403239982844, + "loss": 2.4263, "step": 1244 }, { - "epoch": 0.3987189751801441, - "grad_norm": 0.69140625, - "learning_rate": 3.760478066492971e-05, - "loss": 2.2806, + "epoch": 0.4837182195672762, + "grad_norm": 0.8203125, + "learning_rate": 0.00012293196401459494, + "loss": 2.4282, "step": 1245 }, { - "epoch": 0.3990392313851081, - "grad_norm": 0.54296875, - "learning_rate": 3.758062669920354e-05, - "loss": 2.312, + "epoch": 0.4841067482576917, + "grad_norm": 0.78515625, + "learning_rate": 0.0001227998533976249, + "loss": 2.3939, "step": 1246 }, { - "epoch": 0.39935948759007206, - "grad_norm": 0.33203125, - "learning_rate": 3.7556456997326975e-05, - "loss": 2.1427, + "epoch": 0.4844952769481071, + "grad_norm": 0.84765625, + "learning_rate": 0.00012266770079221457, + "loss": 2.4431, "step": 1247 }, { - "epoch": 0.399679743795036, - "grad_norm": 0.53515625, - "learning_rate": 3.7532271589532065e-05, - "loss": 2.2062, + "epoch": 0.4848838056385226, + "grad_norm": 0.765625, + "learning_rate": 0.00012253550644173754, + "loss": 2.3905, "step": 1248 }, { - "epoch": 0.4, - "grad_norm": 0.470703125, - "learning_rate": 3.75080705060705e-05, - "loss": 2.4721, + "epoch": 0.4852723343289381, + "grad_norm": 0.78125, + "learning_rate": 0.00012240327058964424, + "loss": 2.372, "step": 1249 }, { - "epoch": 0.400320256204964, - "grad_norm": 0.49609375, - "learning_rate": 3.748385377721356e-05, - "loss": 2.5967, + "epoch": 0.4856608630193536, + "grad_norm": 0.76953125, + "learning_rate": 0.00012227099347946155, + "loss": 2.4248, "step": 1250 }, { - "epoch": 0.40064051240992793, - "grad_norm": 0.5859375, - "learning_rate": 3.7459621433252135e-05, - "loss": 2.1675, + "epoch": 0.4860493917097691, + "grad_norm": 0.734375, + "learning_rate": 0.00012213867535479234, + "loss": 2.3693, "step": 1251 }, { - "epoch": 0.40096076861489194, - "grad_norm": 0.42578125, - "learning_rate": 3.74353735044966e-05, - "loss": 1.9072, + "epoch": 0.4864379204001846, + "grad_norm": 0.7109375, + "learning_rate": 0.000122006316459315, + "loss": 2.3079, "step": 1252 }, { - "epoch": 0.4012810248198559, - "grad_norm": 0.388671875, - "learning_rate": 3.741111002127688e-05, - "loss": 1.6543, + "epoch": 0.4868264490906, + "grad_norm": 0.80078125, + "learning_rate": 0.00012187391703678301, + "loss": 2.3852, "step": 1253 }, { - "epoch": 0.40160128102481985, + "epoch": 0.4872149777810155, "grad_norm": 0.75390625, - "learning_rate": 3.738683101394228e-05, - "loss": 1.8961, + "learning_rate": 0.00012174147733102448, + "loss": 2.409, "step": 1254 }, { - "epoch": 0.40192153722978385, - "grad_norm": 0.6328125, - "learning_rate": 3.73625365128616e-05, - "loss": 2.2844, + "epoch": 0.487603506471431, + "grad_norm": 0.84765625, + "learning_rate": 0.00012160899758594176, + "loss": 2.3169, "step": 1255 }, { - "epoch": 0.4022417934347478, - "grad_norm": 1.0703125, - "learning_rate": 3.733822654842298e-05, - "loss": 2.2592, + "epoch": 0.48799203516184647, + "grad_norm": 0.78125, + "learning_rate": 0.00012147647804551078, + "loss": 2.4159, "step": 1256 }, { - "epoch": 0.40256204963971176, - "grad_norm": 0.57421875, - "learning_rate": 3.73139011510339e-05, - "loss": 2.8168, + "epoch": 0.488380563852262, + "grad_norm": 0.765625, + "learning_rate": 0.00012134391895378097, + "loss": 2.3722, "step": 1257 }, { - "epoch": 0.4028823058446757, - "grad_norm": 0.71875, - "learning_rate": 3.728956035112115e-05, - "loss": 1.9843, + "epoch": 0.4887690925426775, + "grad_norm": 0.8203125, + "learning_rate": 0.00012121132055487441, + "loss": 2.3752, "step": 1258 }, { - "epoch": 0.4032025620496397, - "grad_norm": 0.56640625, - "learning_rate": 3.72652041791308e-05, - "loss": 2.2555, + "epoch": 0.4891576212330929, + "grad_norm": 0.796875, + "learning_rate": 0.00012107868309298574, + "loss": 2.3253, "step": 1259 }, { - "epoch": 0.4035228182546037, - "grad_norm": 0.416015625, - "learning_rate": 3.7240832665528124e-05, - "loss": 2.3078, + "epoch": 0.4895461499235084, + "grad_norm": 0.8046875, + "learning_rate": 0.00012094600681238135, + "loss": 2.3405, "step": 1260 }, { - "epoch": 0.40384307445956763, - "grad_norm": 0.53125, - "learning_rate": 3.721644584079761e-05, - "loss": 1.7157, + "epoch": 0.4899346786139239, + "grad_norm": 0.765625, + "learning_rate": 0.00012081329195739928, + "loss": 2.3236, "step": 1261 }, { - "epoch": 0.40416333066453164, - "grad_norm": 0.7421875, - "learning_rate": 3.7192043735442865e-05, - "loss": 2.4214, + "epoch": 0.49032320730433937, + "grad_norm": 0.6953125, + "learning_rate": 0.00012068053877244853, + "loss": 2.408, "step": 1262 }, { - "epoch": 0.4044835868694956, - "grad_norm": 0.5859375, - "learning_rate": 3.7167626379986655e-05, - "loss": 2.5919, + "epoch": 0.49071173599475487, + "grad_norm": 0.69921875, + "learning_rate": 0.0001205477475020087, + "loss": 2.3682, "step": 1263 }, { - "epoch": 0.40480384307445955, - "grad_norm": 0.33984375, - "learning_rate": 3.714319380497077e-05, - "loss": 2.6869, + "epoch": 0.4911002646851703, + "grad_norm": 0.90625, + "learning_rate": 0.0001204149183906296, + "loss": 2.4261, "step": 1264 }, { - "epoch": 0.40512409927942356, - "grad_norm": 0.77734375, - "learning_rate": 3.7118746040956076e-05, - "loss": 2.0477, + "epoch": 0.4914887933755858, + "grad_norm": 0.85546875, + "learning_rate": 0.00012028205168293056, + "loss": 2.3513, "step": 1265 }, { - "epoch": 0.4054443554843875, - "grad_norm": 0.5078125, - "learning_rate": 3.709428311852242e-05, - "loss": 1.8267, + "epoch": 0.4918773220660013, + "grad_norm": 0.70703125, + "learning_rate": 0.0001201491476236003, + "loss": 2.4025, "step": 1266 }, { - "epoch": 0.40576461168935146, - "grad_norm": 0.52734375, - "learning_rate": 3.706980506826863e-05, - "loss": 2.1828, + "epoch": 0.49226585075641677, + "grad_norm": 0.8046875, + "learning_rate": 0.00012001620645739629, + "loss": 2.3808, "step": 1267 }, { - "epoch": 0.40608486789431547, - "grad_norm": 0.515625, - "learning_rate": 3.704531192081241e-05, - "loss": 2.2251, + "epoch": 0.49265437944683227, + "grad_norm": 0.76953125, + "learning_rate": 0.0001198832284291443, + "loss": 2.4421, "step": 1268 }, { - "epoch": 0.4064051240992794, - "grad_norm": 0.369140625, - "learning_rate": 3.70208037067904e-05, - "loss": 1.7606, + "epoch": 0.49304290813724777, + "grad_norm": 0.81640625, + "learning_rate": 0.00011975021378373802, + "loss": 2.4174, "step": 1269 }, { - "epoch": 0.4067253803042434, - "grad_norm": 0.494140625, - "learning_rate": 3.699628045685805e-05, - "loss": 2.4729, + "epoch": 0.4934314368276632, + "grad_norm": 0.71875, + "learning_rate": 0.00011961716276613857, + "loss": 2.3233, "step": 1270 }, { - "epoch": 0.4070456365092074, - "grad_norm": 0.40625, - "learning_rate": 3.697174220168965e-05, - "loss": 1.5516, + "epoch": 0.4938199655180787, + "grad_norm": 0.73046875, + "learning_rate": 0.00011948407562137404, + "loss": 2.4112, "step": 1271 }, { - "epoch": 0.40736589271417134, - "grad_norm": 0.474609375, - "learning_rate": 3.694718897197822e-05, - "loss": 2.0469, + "epoch": 0.4942084942084942, + "grad_norm": 0.7734375, + "learning_rate": 0.0001193509525945391, + "loss": 2.4136, "step": 1272 }, { - "epoch": 0.4076861489191353, - "grad_norm": 0.486328125, - "learning_rate": 3.692262079843553e-05, - "loss": 1.997, + "epoch": 0.49459702289890967, + "grad_norm": 0.72265625, + "learning_rate": 0.00011921779393079438, + "loss": 2.3525, "step": 1273 }, { - "epoch": 0.4080064051240993, - "grad_norm": 0.48828125, - "learning_rate": 3.689803771179207e-05, - "loss": 2.3182, + "epoch": 0.49498555158932517, + "grad_norm": 0.734375, + "learning_rate": 0.00011908459987536629, + "loss": 2.3849, "step": 1274 }, { - "epoch": 0.40832666132906326, - "grad_norm": 0.474609375, - "learning_rate": 3.687343974279691e-05, - "loss": 2.4617, + "epoch": 0.49537408027974067, + "grad_norm": 0.74609375, + "learning_rate": 0.0001189513706735463, + "loss": 2.4242, "step": 1275 }, { - "epoch": 0.4086469175340272, - "grad_norm": 0.609375, - "learning_rate": 3.684882692221782e-05, - "loss": 2.3608, + "epoch": 0.4957626089701561, + "grad_norm": 0.73046875, + "learning_rate": 0.00011881810657069068, + "loss": 2.3559, "step": 1276 }, { - "epoch": 0.4089671737389912, - "grad_norm": 0.458984375, - "learning_rate": 3.682419928084109e-05, - "loss": 2.1872, + "epoch": 0.4961511376605716, + "grad_norm": 0.7734375, + "learning_rate": 0.00011868480781221995, + "loss": 2.3649, "step": 1277 }, { - "epoch": 0.40928742994395517, - "grad_norm": 0.5703125, - "learning_rate": 3.679955684947158e-05, - "loss": 2.0972, + "epoch": 0.4965396663509871, + "grad_norm": 0.73828125, + "learning_rate": 0.00011855147464361845, + "loss": 2.2891, "step": 1278 }, { - "epoch": 0.4096076861489191, - "grad_norm": 0.40625, - "learning_rate": 3.6774899658932615e-05, - "loss": 2.086, + "epoch": 0.49692819504140257, + "grad_norm": 0.73828125, + "learning_rate": 0.00011841810731043386, + "loss": 2.4437, "step": 1279 }, { - "epoch": 0.4099279423538831, - "grad_norm": 0.4765625, - "learning_rate": 3.6750227740066015e-05, - "loss": 2.3214, + "epoch": 0.49731672373181807, + "grad_norm": 0.78125, + "learning_rate": 0.00011828470605827682, + "loss": 2.3526, "step": 1280 }, { - "epoch": 0.4102481985588471, - "grad_norm": 1.6640625, - "learning_rate": 3.6725541123732e-05, - "loss": 2.2449, + "epoch": 0.49770525242223357, + "grad_norm": 0.78515625, + "learning_rate": 0.00011815127113282047, + "loss": 2.3752, "step": 1281 }, { - "epoch": 0.41056845476381104, - "grad_norm": 3.421875, - "learning_rate": 3.6700839840809206e-05, - "loss": 1.619, + "epoch": 0.498093781112649, + "grad_norm": 0.73046875, + "learning_rate": 0.00011801780277979986, + "loss": 2.3949, "step": 1282 }, { - "epoch": 0.410888710968775, - "grad_norm": 0.56640625, - "learning_rate": 3.6676123922194574e-05, - "loss": 2.1188, + "epoch": 0.4984823098030645, + "grad_norm": 0.75390625, + "learning_rate": 0.00011788430124501167, + "loss": 2.4448, "step": 1283 }, { - "epoch": 0.411208967173739, - "grad_norm": 0.5078125, - "learning_rate": 3.6651393398803366e-05, - "loss": 2.6604, + "epoch": 0.49887083849348, + "grad_norm": 0.7578125, + "learning_rate": 0.0001177507667743137, + "loss": 2.3877, "step": 1284 }, { - "epoch": 0.41152922337870296, - "grad_norm": 1.15625, - "learning_rate": 3.6626648301569124e-05, - "loss": 1.9312, + "epoch": 0.49925936718389546, + "grad_norm": 0.8359375, + "learning_rate": 0.00011761719961362434, + "loss": 2.3495, "step": 1285 }, { - "epoch": 0.4118494795836669, - "grad_norm": 0.474609375, - "learning_rate": 3.6601888661443603e-05, - "loss": 2.0071, + "epoch": 0.49964789587431097, + "grad_norm": 0.85546875, + "learning_rate": 0.00011748360000892227, + "loss": 2.4588, "step": 1286 }, { - "epoch": 0.4121697357886309, - "grad_norm": 0.66015625, - "learning_rate": 3.657711450939676e-05, - "loss": 2.1451, + "epoch": 0.5000364245647264, + "grad_norm": 0.734375, + "learning_rate": 0.00011734996820624581, + "loss": 2.3902, "step": 1287 }, { - "epoch": 0.4124899919935949, - "grad_norm": 0.61328125, - "learning_rate": 3.6552325876416704e-05, - "loss": 2.6817, + "epoch": 0.5004249532551419, + "grad_norm": 0.76171875, + "learning_rate": 0.0001172163044516927, + "loss": 2.3359, "step": 1288 }, { - "epoch": 0.4128102481985588, - "grad_norm": 0.37890625, - "learning_rate": 3.652752279350965e-05, - "loss": 2.2703, + "epoch": 0.5008134819455574, + "grad_norm": 0.70703125, + "learning_rate": 0.00011708260899141943, + "loss": 2.3574, "step": 1289 }, { - "epoch": 0.41313050440352284, - "grad_norm": 0.5078125, - "learning_rate": 3.65027052916999e-05, - "loss": 1.6399, + "epoch": 0.5012020106359729, + "grad_norm": 0.703125, + "learning_rate": 0.00011694888207164091, + "loss": 2.3206, "step": 1290 }, { - "epoch": 0.4134507606084868, - "grad_norm": 0.478515625, - "learning_rate": 3.647787340202975e-05, - "loss": 2.139, + "epoch": 0.5015905393263884, + "grad_norm": 0.7578125, + "learning_rate": 0.00011681512393862999, + "loss": 2.4011, "step": 1291 }, { - "epoch": 0.41377101681345074, - "grad_norm": 0.67578125, - "learning_rate": 3.6453027155559565e-05, - "loss": 2.0271, + "epoch": 0.5019790680168039, + "grad_norm": 0.69140625, + "learning_rate": 0.00011668133483871699, + "loss": 2.3868, "step": 1292 }, { - "epoch": 0.41409127301841475, - "grad_norm": 0.48828125, - "learning_rate": 3.6428166583367615e-05, - "loss": 2.1767, + "epoch": 0.5023675967072193, + "grad_norm": 0.73828125, + "learning_rate": 0.00011654751501828927, + "loss": 2.373, "step": 1293 }, { - "epoch": 0.4144115292233787, - "grad_norm": 1.1796875, - "learning_rate": 3.640329171655008e-05, - "loss": 1.9477, + "epoch": 0.5027561253976348, + "grad_norm": 0.71875, + "learning_rate": 0.00011641366472379078, + "loss": 2.3772, "step": 1294 }, { - "epoch": 0.41473178542834266, - "grad_norm": 0.49609375, - "learning_rate": 3.637840258622106e-05, - "loss": 3.0508, + "epoch": 0.5031446540880503, + "grad_norm": 0.77734375, + "learning_rate": 0.00011627978420172156, + "loss": 2.4444, "step": 1295 }, { - "epoch": 0.41505204163330667, - "grad_norm": 0.455078125, - "learning_rate": 3.635349922351249e-05, - "loss": 2.6344, + "epoch": 0.5035331827784658, + "grad_norm": 0.71875, + "learning_rate": 0.00011614587369863737, + "loss": 2.3138, "step": 1296 }, { - "epoch": 0.4153722978382706, - "grad_norm": 4.125, - "learning_rate": 3.632858165957407e-05, - "loss": 2.1484, + "epoch": 0.5039217114688813, + "grad_norm": 0.6953125, + "learning_rate": 0.00011601193346114905, + "loss": 2.2997, "step": 1297 }, { - "epoch": 0.4156925540432346, - "grad_norm": 0.3984375, - "learning_rate": 3.63036499255733e-05, - "loss": 2.0567, + "epoch": 0.5043102401592967, + "grad_norm": 0.671875, + "learning_rate": 0.00011587796373592237, + "loss": 2.3125, "step": 1298 }, { - "epoch": 0.4160128102481986, - "grad_norm": 0.890625, - "learning_rate": 3.627870405269539e-05, - "loss": 2.6683, + "epoch": 0.5046987688497122, + "grad_norm": 0.73046875, + "learning_rate": 0.00011574396476967732, + "loss": 2.423, "step": 1299 }, { - "epoch": 0.41633306645316254, - "grad_norm": 0.55859375, - "learning_rate": 3.625374407214325e-05, - "loss": 2.0889, + "epoch": 0.5050872975401277, + "grad_norm": 0.703125, + "learning_rate": 0.00011560993680918774, + "loss": 2.3215, "step": 1300 }, { - "epoch": 0.4166533226581265, - "grad_norm": 0.5625, - "learning_rate": 3.62287700151374e-05, - "loss": 2.3849, + "epoch": 0.5054758262305432, + "grad_norm": 0.72265625, + "learning_rate": 0.00011547588010128088, + "loss": 2.3576, "step": 1301 }, { - "epoch": 0.4169735788630905, - "grad_norm": 0.46875, - "learning_rate": 3.6203781912916e-05, - "loss": 2.5107, + "epoch": 0.5058643549209587, + "grad_norm": 0.76953125, + "learning_rate": 0.0001153417948928369, + "loss": 2.3201, "step": 1302 }, { - "epoch": 0.41729383506805445, - "grad_norm": 0.55078125, - "learning_rate": 3.6178779796734785e-05, - "loss": 2.0597, + "epoch": 0.5062528836113742, + "grad_norm": 0.765625, + "learning_rate": 0.00011520768143078853, + "loss": 2.4282, "step": 1303 }, { - "epoch": 0.4176140912730184, - "grad_norm": 0.70703125, - "learning_rate": 3.615376369786699e-05, - "loss": 2.3948, + "epoch": 0.5066414123017896, + "grad_norm": 0.77734375, + "learning_rate": 0.00011507353996212043, + "loss": 2.3862, "step": 1304 }, { - "epoch": 0.41793434747798236, - "grad_norm": 0.53125, - "learning_rate": 3.6128733647603345e-05, - "loss": 2.1924, + "epoch": 0.5070299409922051, + "grad_norm": 0.76953125, + "learning_rate": 0.00011493937073386889, + "loss": 2.3773, "step": 1305 }, { - "epoch": 0.41825460368294637, - "grad_norm": 0.640625, - "learning_rate": 3.610368967725205e-05, - "loss": 2.1514, + "epoch": 0.5074184696826206, + "grad_norm": 0.7421875, + "learning_rate": 0.00011480517399312134, + "loss": 2.4023, "step": 1306 }, { - "epoch": 0.4185748598879103, - "grad_norm": 0.59375, - "learning_rate": 3.607863181813871e-05, - "loss": 2.1708, + "epoch": 0.5078069983730361, + "grad_norm": 0.73828125, + "learning_rate": 0.00011467094998701582, + "loss": 2.3434, "step": 1307 }, { - "epoch": 0.4188951160928743, - "grad_norm": 0.5234375, - "learning_rate": 3.6053560101606285e-05, - "loss": 2.4148, + "epoch": 0.5081955270634516, + "grad_norm": 0.75, + "learning_rate": 0.00011453669896274066, + "loss": 2.3864, "step": 1308 }, { - "epoch": 0.4192153722978383, - "grad_norm": 0.486328125, - "learning_rate": 3.602847455901509e-05, - "loss": 2.3323, + "epoch": 0.5085840557538671, + "grad_norm": 0.765625, + "learning_rate": 0.00011440242116753384, + "loss": 2.3932, "step": 1309 }, { - "epoch": 0.41953562850280224, - "grad_norm": 0.68359375, - "learning_rate": 3.600337522174272e-05, - "loss": 2.2687, + "epoch": 0.5089725844442825, + "grad_norm": 0.71875, + "learning_rate": 0.00011426811684868275, + "loss": 2.3176, "step": 1310 }, { - "epoch": 0.4198558847077662, - "grad_norm": 0.45703125, - "learning_rate": 3.597826212118404e-05, - "loss": 2.5078, + "epoch": 0.509361113134698, + "grad_norm": 0.76171875, + "learning_rate": 0.00011413378625352358, + "loss": 2.4461, "step": 1311 }, { - "epoch": 0.4201761409127302, - "grad_norm": 0.609375, - "learning_rate": 3.5953135288751094e-05, - "loss": 2.1704, + "epoch": 0.5097496418251135, + "grad_norm": 0.7265625, + "learning_rate": 0.0001139994296294409, + "loss": 2.3435, "step": 1312 }, { - "epoch": 0.42049639711769415, - "grad_norm": 0.5078125, - "learning_rate": 3.5927994755873157e-05, - "loss": 2.3606, + "epoch": 0.510138170515529, + "grad_norm": 0.7265625, + "learning_rate": 0.0001138650472238672, + "loss": 2.329, "step": 1313 }, { - "epoch": 0.4208166533226581, - "grad_norm": 0.70703125, - "learning_rate": 3.5902840553996594e-05, - "loss": 1.9514, + "epoch": 0.5105266992059445, + "grad_norm": 0.69140625, + "learning_rate": 0.00011373063928428252, + "loss": 2.3996, "step": 1314 }, { - "epoch": 0.4211369095276221, - "grad_norm": 0.455078125, - "learning_rate": 3.58776727145849e-05, - "loss": 2.6369, + "epoch": 0.51091522789636, + "grad_norm": 0.74609375, + "learning_rate": 0.00011359620605821384, + "loss": 2.4047, "step": 1315 }, { - "epoch": 0.42145716573258607, - "grad_norm": 0.421875, - "learning_rate": 3.58524912691186e-05, - "loss": 2.2219, + "epoch": 0.5113037565867754, + "grad_norm": 0.71875, + "learning_rate": 0.0001134617477932348, + "loss": 2.4164, "step": 1316 }, { - "epoch": 0.42177742193755, - "grad_norm": 0.4921875, - "learning_rate": 3.5827296249095264e-05, - "loss": 1.7456, + "epoch": 0.5116922852771909, + "grad_norm": 0.73828125, + "learning_rate": 0.00011332726473696505, + "loss": 2.326, "step": 1317 }, { - "epoch": 0.42209767814251403, - "grad_norm": 0.5078125, - "learning_rate": 3.580208768602941e-05, - "loss": 2.2512, + "epoch": 0.5120808139676064, + "grad_norm": 0.70703125, + "learning_rate": 0.00011319275713706996, + "loss": 2.3532, "step": 1318 }, { - "epoch": 0.422417934347478, - "grad_norm": 0.6484375, - "learning_rate": 3.577686561145254e-05, - "loss": 2.1483, + "epoch": 0.5124693426580219, + "grad_norm": 0.73828125, + "learning_rate": 0.00011305822524126007, + "loss": 2.3886, "step": 1319 }, { - "epoch": 0.42273819055244194, - "grad_norm": 0.3671875, - "learning_rate": 3.575163005691302e-05, - "loss": 1.9452, + "epoch": 0.5128578713484374, + "grad_norm": 0.7421875, + "learning_rate": 0.0001129236692972907, + "loss": 2.4516, "step": 1320 }, { - "epoch": 0.42305844675740595, - "grad_norm": 0.65625, - "learning_rate": 3.572638105397607e-05, - "loss": 2.1301, + "epoch": 0.5132464000388529, + "grad_norm": 0.7265625, + "learning_rate": 0.00011278908955296143, + "loss": 2.371, "step": 1321 }, { - "epoch": 0.4233787029623699, - "grad_norm": 0.431640625, - "learning_rate": 3.5701118634223785e-05, - "loss": 2.099, + "epoch": 0.5136349287292683, + "grad_norm": 0.71875, + "learning_rate": 0.00011265448625611568, + "loss": 2.3791, "step": 1322 }, { - "epoch": 0.42369895916733386, - "grad_norm": 0.45703125, - "learning_rate": 3.567584282925498e-05, - "loss": 2.2, + "epoch": 0.5140234574196838, + "grad_norm": 0.69140625, + "learning_rate": 0.00011251985965464022, + "loss": 2.4764, "step": 1323 }, { - "epoch": 0.42401921537229786, - "grad_norm": 0.466796875, - "learning_rate": 3.565055367068525e-05, - "loss": 2.4591, + "epoch": 0.5144119861100993, + "grad_norm": 0.6953125, + "learning_rate": 0.00011238520999646479, + "loss": 2.4049, "step": 1324 }, { - "epoch": 0.4243394715772618, - "grad_norm": 0.52734375, - "learning_rate": 3.562525119014688e-05, - "loss": 2.1612, + "epoch": 0.5148005148005148, + "grad_norm": 0.76171875, + "learning_rate": 0.00011225053752956152, + "loss": 2.3209, "step": 1325 }, { - "epoch": 0.42465972778222577, - "grad_norm": 0.5, - "learning_rate": 3.5599935419288824e-05, - "loss": 1.9581, + "epoch": 0.5151890434909303, + "grad_norm": 0.6796875, + "learning_rate": 0.00011211584250194463, + "loss": 2.3245, "step": 1326 }, { - "epoch": 0.4249799839871897, - "grad_norm": 0.54296875, - "learning_rate": 3.557460638977666e-05, - "loss": 2.4488, + "epoch": 0.5155775721813458, + "grad_norm": 0.77734375, + "learning_rate": 0.0001119811251616698, + "loss": 2.3636, "step": 1327 }, { - "epoch": 0.42530024019215373, - "grad_norm": 0.5234375, - "learning_rate": 3.554926413329254e-05, - "loss": 2.6207, + "epoch": 0.5159661008717612, + "grad_norm": 0.69140625, + "learning_rate": 0.00011184638575683388, + "loss": 2.3771, "step": 1328 }, { - "epoch": 0.4256204963971177, - "grad_norm": 0.494140625, - "learning_rate": 3.552390868153516e-05, - "loss": 2.0293, + "epoch": 0.5163546295621767, + "grad_norm": 0.75390625, + "learning_rate": 0.00011171162453557431, + "loss": 2.4129, "step": 1329 }, { - "epoch": 0.42594075260208164, - "grad_norm": 0.390625, - "learning_rate": 3.549854006621976e-05, - "loss": 2.2195, + "epoch": 0.5167431582525922, + "grad_norm": 0.76171875, + "learning_rate": 0.00011157684174606872, + "loss": 2.3208, "step": 1330 }, { - "epoch": 0.42626100880704565, - "grad_norm": 0.99609375, - "learning_rate": 3.547315831907798e-05, - "loss": 2.9019, + "epoch": 0.5171316869430077, + "grad_norm": 0.703125, + "learning_rate": 0.00011144203763653443, + "loss": 2.4374, "step": 1331 }, { - "epoch": 0.4265812650120096, - "grad_norm": 0.515625, - "learning_rate": 3.544776347185793e-05, - "loss": 2.5745, + "epoch": 0.5175202156334232, + "grad_norm": 0.74609375, + "learning_rate": 0.00011130721245522808, + "loss": 2.4326, "step": 1332 }, { - "epoch": 0.42690152121697356, - "grad_norm": 0.54296875, - "learning_rate": 3.54223555563241e-05, - "loss": 2.6656, + "epoch": 0.5179087443238387, + "grad_norm": 0.69921875, + "learning_rate": 0.00011117236645044506, + "loss": 2.3765, "step": 1333 }, { - "epoch": 0.42722177742193757, - "grad_norm": 0.44921875, - "learning_rate": 3.5396934604257326e-05, - "loss": 2.2501, + "epoch": 0.5182972730142541, + "grad_norm": 0.6953125, + "learning_rate": 0.00011103749987051916, + "loss": 2.3262, "step": 1334 }, { - "epoch": 0.4275420336269015, - "grad_norm": 0.4609375, - "learning_rate": 3.537150064745474e-05, - "loss": 2.2139, + "epoch": 0.5186858017046696, + "grad_norm": 0.703125, + "learning_rate": 0.00011090261296382202, + "loss": 2.3944, "step": 1335 }, { - "epoch": 0.42786228983186547, - "grad_norm": 0.44921875, - "learning_rate": 3.534605371772974e-05, - "loss": 1.9124, + "epoch": 0.5190743303950851, + "grad_norm": 0.6796875, + "learning_rate": 0.00011076770597876272, + "loss": 2.3092, "step": 1336 }, { - "epoch": 0.4281825460368295, - "grad_norm": 0.56640625, - "learning_rate": 3.5320593846911973e-05, - "loss": 2.6009, + "epoch": 0.5194628590855006, + "grad_norm": 0.70703125, + "learning_rate": 0.00011063277916378736, + "loss": 2.3485, "step": 1337 }, { - "epoch": 0.42850280224179343, - "grad_norm": 0.41796875, - "learning_rate": 3.529512106684724e-05, - "loss": 1.5908, + "epoch": 0.5198513877759161, + "grad_norm": 0.765625, + "learning_rate": 0.0001104978327673785, + "loss": 2.4069, "step": 1338 }, { - "epoch": 0.4288230584467574, - "grad_norm": 0.41015625, - "learning_rate": 3.526963540939752e-05, - "loss": 2.011, - "step": 1339 + "epoch": 0.5202399164663316, + "grad_norm": 0.7421875, + "learning_rate": 0.00011036286703805478, + "loss": 2.4546, + "step": 1339 }, { - "epoch": 0.4291433146517214, - "grad_norm": 0.54296875, - "learning_rate": 3.5244136906440886e-05, - "loss": 2.0686, + "epoch": 0.520628445156747, + "grad_norm": 0.71875, + "learning_rate": 0.00011022788222437048, + "loss": 2.3471, "step": 1340 }, { - "epoch": 0.42946357085668535, - "grad_norm": 0.9453125, - "learning_rate": 3.521862558987148e-05, - "loss": 1.9993, + "epoch": 0.5210169738471625, + "grad_norm": 0.71875, + "learning_rate": 0.00011009287857491497, + "loss": 2.4365, "step": 1341 }, { - "epoch": 0.4297838270616493, - "grad_norm": 0.419921875, - "learning_rate": 3.5193101491599453e-05, - "loss": 2.6751, + "epoch": 0.521405502537578, + "grad_norm": 0.6796875, + "learning_rate": 0.00010995785633831233, + "loss": 2.358, "step": 1342 }, { - "epoch": 0.4301040832666133, - "grad_norm": 0.453125, - "learning_rate": 3.5167564643550975e-05, - "loss": 2.1176, + "epoch": 0.5217940312279935, + "grad_norm": 0.73828125, + "learning_rate": 0.00010982281576322091, + "loss": 2.3183, "step": 1343 }, { - "epoch": 0.43042433947157727, - "grad_norm": 0.44140625, - "learning_rate": 3.5142015077668155e-05, - "loss": 2.3972, + "epoch": 0.522182559918409, + "grad_norm": 0.72265625, + "learning_rate": 0.00010968775709833274, + "loss": 2.4138, "step": 1344 }, { - "epoch": 0.4307445956765412, - "grad_norm": 0.291015625, - "learning_rate": 3.5116452825909e-05, - "loss": 1.8073, + "epoch": 0.5225710886088245, + "grad_norm": 0.73046875, + "learning_rate": 0.00010955268059237327, + "loss": 2.3831, "step": 1345 }, { - "epoch": 0.43106485188150523, - "grad_norm": 0.62890625, - "learning_rate": 3.5090877920247375e-05, - "loss": 2.6247, + "epoch": 0.5229596172992399, + "grad_norm": 0.765625, + "learning_rate": 0.00010941758649410075, + "loss": 2.3526, "step": 1346 }, { - "epoch": 0.4313851080864692, - "grad_norm": 0.41796875, - "learning_rate": 3.5065290392673e-05, - "loss": 1.8295, + "epoch": 0.5233481459896554, + "grad_norm": 0.7265625, + "learning_rate": 0.0001092824750523058, + "loss": 2.3985, "step": 1347 }, { - "epoch": 0.43170536429143314, - "grad_norm": 0.50390625, - "learning_rate": 3.503969027519137e-05, - "loss": 2.4511, + "epoch": 0.5237366746800709, + "grad_norm": 0.73046875, + "learning_rate": 0.00010914734651581108, + "loss": 2.4037, "step": 1348 }, { - "epoch": 0.43202562049639714, - "grad_norm": 0.474609375, - "learning_rate": 3.501407759982372e-05, - "loss": 2.578, + "epoch": 0.5241252033704864, + "grad_norm": 0.7578125, + "learning_rate": 0.00010901220113347062, + "loss": 2.4123, "step": 1349 }, { - "epoch": 0.4323458767013611, - "grad_norm": 0.64453125, - "learning_rate": 3.498845239860702e-05, - "loss": 2.1768, + "epoch": 0.5245137320609019, + "grad_norm": 0.7578125, + "learning_rate": 0.00010887703915416951, + "loss": 2.3917, "step": 1350 }, { - "epoch": 0.43266613290632505, - "grad_norm": 0.3984375, - "learning_rate": 3.496281470359386e-05, - "loss": 2.084, + "epoch": 0.5249022607513174, + "grad_norm": 0.765625, + "learning_rate": 0.00010874186082682345, + "loss": 2.4087, "step": 1351 }, { - "epoch": 0.432986389111289, - "grad_norm": 0.423828125, - "learning_rate": 3.493716454685251e-05, - "loss": 2.215, + "epoch": 0.5252907894417328, + "grad_norm": 0.69140625, + "learning_rate": 0.00010860666640037825, + "loss": 2.3286, "step": 1352 }, { - "epoch": 0.433306645316253, - "grad_norm": 0.64453125, - "learning_rate": 3.491150196046679e-05, - "loss": 2.1564, + "epoch": 0.5256793181321483, + "grad_norm": 0.74609375, + "learning_rate": 0.00010847145612380923, + "loss": 2.4452, "step": 1353 }, { - "epoch": 0.43362690152121697, - "grad_norm": 0.60546875, - "learning_rate": 3.4885826976536074e-05, - "loss": 2.0137, + "epoch": 0.5260678468225638, + "grad_norm": 0.82421875, + "learning_rate": 0.00010833623024612105, + "loss": 2.3288, "step": 1354 }, { - "epoch": 0.4339471577261809, - "grad_norm": 0.486328125, - "learning_rate": 3.486013962717526e-05, - "loss": 1.9845, + "epoch": 0.5264563755129793, + "grad_norm": 0.76953125, + "learning_rate": 0.00010820098901634705, + "loss": 2.3248, "step": 1355 }, { - "epoch": 0.43426741393114493, - "grad_norm": 0.357421875, - "learning_rate": 3.483443994451471e-05, - "loss": 1.7616, + "epoch": 0.5268449042033948, + "grad_norm": 0.6953125, + "learning_rate": 0.0001080657326835489, + "loss": 2.3524, "step": 1356 }, { - "epoch": 0.4345876701361089, - "grad_norm": 0.439453125, - "learning_rate": 3.480872796070018e-05, - "loss": 2.5005, + "epoch": 0.5272334328938103, + "grad_norm": 0.73046875, + "learning_rate": 0.00010793046149681593, + "loss": 2.3483, "step": 1357 }, { - "epoch": 0.43490792634107284, - "grad_norm": 0.46484375, - "learning_rate": 3.478300370789286e-05, - "loss": 2.4385, + "epoch": 0.5276219615842257, + "grad_norm": 0.71875, + "learning_rate": 0.00010779517570526499, + "loss": 2.4338, "step": 1358 }, { - "epoch": 0.43522818254603685, - "grad_norm": 0.52734375, - "learning_rate": 3.475726721826926e-05, - "loss": 2.3023, + "epoch": 0.5280104902746412, + "grad_norm": 0.7109375, + "learning_rate": 0.00010765987555803973, + "loss": 2.4, "step": 1359 }, { - "epoch": 0.4355484387510008, - "grad_norm": 0.48046875, - "learning_rate": 3.473151852402119e-05, - "loss": 2.0413, + "epoch": 0.5283990189650567, + "grad_norm": 0.7578125, + "learning_rate": 0.00010752456130431032, + "loss": 2.3773, "step": 1360 }, { - "epoch": 0.43586869495596475, - "grad_norm": 0.62109375, - "learning_rate": 3.470575765735574e-05, - "loss": 2.4551, + "epoch": 0.5287875476554722, + "grad_norm": 0.734375, + "learning_rate": 0.00010738923319327281, + "loss": 2.333, "step": 1361 }, { - "epoch": 0.43618895116092876, - "grad_norm": 0.50390625, - "learning_rate": 3.4679984650495225e-05, - "loss": 2.5011, + "epoch": 0.5291760763458877, + "grad_norm": 0.72265625, + "learning_rate": 0.0001072538914741488, + "loss": 2.4401, "step": 1362 }, { - "epoch": 0.4365092073658927, - "grad_norm": 0.52734375, - "learning_rate": 3.465419953567714e-05, - "loss": 1.651, + "epoch": 0.5295646050363032, + "grad_norm": 0.73046875, + "learning_rate": 0.00010711853639618497, + "loss": 2.3724, "step": 1363 }, { - "epoch": 0.43682946357085667, - "grad_norm": 0.54296875, - "learning_rate": 3.4628402345154136e-05, - "loss": 2.2163, + "epoch": 0.5299531337267186, + "grad_norm": 0.7421875, + "learning_rate": 0.00010698316820865264, + "loss": 2.43, "step": 1364 }, { - "epoch": 0.4371497197758207, - "grad_norm": 0.42578125, - "learning_rate": 3.4602593111193946e-05, - "loss": 2.1722, + "epoch": 0.5303416624171341, + "grad_norm": 0.71484375, + "learning_rate": 0.00010684778716084722, + "loss": 2.3521, "step": 1365 }, { - "epoch": 0.43746997598078463, - "grad_norm": 0.455078125, - "learning_rate": 3.457677186607938e-05, - "loss": 2.5205, + "epoch": 0.5307301911075496, + "grad_norm": 0.71875, + "learning_rate": 0.0001067123935020878, + "loss": 2.4063, "step": 1366 }, { - "epoch": 0.4377902321857486, - "grad_norm": 0.66796875, - "learning_rate": 3.455093864210828e-05, - "loss": 2.0087, + "epoch": 0.5311187197979651, + "grad_norm": 0.6875, + "learning_rate": 0.0001065769874817167, + "loss": 2.3809, "step": 1367 }, { - "epoch": 0.4381104883907126, - "grad_norm": 0.65625, - "learning_rate": 3.452509347159346e-05, - "loss": 2.0194, + "epoch": 0.5315072484883806, + "grad_norm": 0.7421875, + "learning_rate": 0.00010644156934909901, + "loss": 2.3949, "step": 1368 }, { - "epoch": 0.43843074459567655, - "grad_norm": 0.48046875, - "learning_rate": 3.449923638686269e-05, - "loss": 2.6472, + "epoch": 0.531895777178796, + "grad_norm": 0.80859375, + "learning_rate": 0.00010630613935362217, + "loss": 2.3989, "step": 1369 }, { - "epoch": 0.4387510008006405, - "grad_norm": 0.419921875, - "learning_rate": 3.447336742025862e-05, - "loss": 2.1153, + "epoch": 0.5322843058692115, + "grad_norm": 0.7109375, + "learning_rate": 0.00010617069774469543, + "loss": 2.4125, "step": 1370 }, { - "epoch": 0.4390712570056045, - "grad_norm": 0.5078125, - "learning_rate": 3.4447486604138776e-05, - "loss": 2.0821, + "epoch": 0.532672834559627, + "grad_norm": 0.703125, + "learning_rate": 0.00010603524477174937, + "loss": 2.3935, "step": 1371 }, { - "epoch": 0.43939151321056846, - "grad_norm": 0.51953125, - "learning_rate": 3.442159397087551e-05, - "loss": 2.3243, + "epoch": 0.5330613632500425, + "grad_norm": 0.74609375, + "learning_rate": 0.00010589978068423561, + "loss": 2.4038, "step": 1372 }, { - "epoch": 0.4397117694155324, - "grad_norm": 0.65625, - "learning_rate": 3.4395689552855955e-05, - "loss": 2.3805, + "epoch": 0.533449891940458, + "grad_norm": 0.6953125, + "learning_rate": 0.00010576430573162612, + "loss": 2.3936, "step": 1373 }, { - "epoch": 0.44003202562049637, - "grad_norm": 0.70703125, - "learning_rate": 3.436977338248197e-05, - "loss": 2.2077, + "epoch": 0.5338384206308735, + "grad_norm": 0.6796875, + "learning_rate": 0.00010562882016341303, + "loss": 2.336, "step": 1374 }, { - "epoch": 0.4403522818254604, - "grad_norm": 0.5703125, - "learning_rate": 3.434384549217013e-05, - "loss": 2.3821, + "epoch": 0.5342269493212889, + "grad_norm": 0.68359375, + "learning_rate": 0.00010549332422910782, + "loss": 2.3207, "step": 1375 }, { - "epoch": 0.44067253803042433, - "grad_norm": 0.9453125, - "learning_rate": 3.431790591435167e-05, - "loss": 2.0472, + "epoch": 0.5346154780117044, + "grad_norm": 0.72265625, + "learning_rate": 0.00010535781817824126, + "loss": 2.3884, "step": 1376 }, { - "epoch": 0.4409927942353883, - "grad_norm": 0.26171875, - "learning_rate": 3.4291954681472427e-05, - "loss": 1.6111, + "epoch": 0.5350040067021199, + "grad_norm": 0.69140625, + "learning_rate": 0.0001052223022603626, + "loss": 2.3599, "step": 1377 }, { - "epoch": 0.4413130504403523, - "grad_norm": 0.423828125, - "learning_rate": 3.426599182599283e-05, - "loss": 2.5023, + "epoch": 0.5353925353925354, + "grad_norm": 0.6875, + "learning_rate": 0.00010508677672503942, + "loss": 2.4099, "step": 1378 }, { - "epoch": 0.44163330664531625, - "grad_norm": 0.42578125, - "learning_rate": 3.424001738038784e-05, - "loss": 2.4245, + "epoch": 0.5357810640829509, + "grad_norm": 0.69921875, + "learning_rate": 0.00010495124182185679, + "loss": 2.4312, "step": 1379 }, { - "epoch": 0.4419535628502802, - "grad_norm": 0.375, - "learning_rate": 3.4214031377146915e-05, - "loss": 2.2083, + "epoch": 0.5361695927733664, + "grad_norm": 0.6875, + "learning_rate": 0.00010481569780041723, + "loss": 2.3607, "step": 1380 }, { - "epoch": 0.4422738190552442, - "grad_norm": 0.80859375, - "learning_rate": 3.4188033848773995e-05, - "loss": 2.5568, + "epoch": 0.5365581214637818, + "grad_norm": 0.71484375, + "learning_rate": 0.00010468014491033995, + "loss": 2.4015, "step": 1381 }, { - "epoch": 0.44259407526020816, - "grad_norm": 0.60546875, - "learning_rate": 3.4162024827787384e-05, - "loss": 2.4897, + "epoch": 0.5369466501541973, + "grad_norm": 0.6796875, + "learning_rate": 0.00010454458340126059, + "loss": 2.3508, "step": 1382 }, { - "epoch": 0.4429143314651721, - "grad_norm": 0.48046875, - "learning_rate": 3.4136004346719815e-05, - "loss": 2.2791, + "epoch": 0.5373351788446128, + "grad_norm": 0.71484375, + "learning_rate": 0.0001044090135228305, + "loss": 2.368, "step": 1383 }, { - "epoch": 0.4432345876701361, - "grad_norm": 0.4296875, - "learning_rate": 3.410997243811832e-05, - "loss": 1.5351, + "epoch": 0.5377237075350283, + "grad_norm": 0.71875, + "learning_rate": 0.00010427343552471657, + "loss": 2.4, "step": 1384 }, { - "epoch": 0.4435548438751001, - "grad_norm": 0.54296875, - "learning_rate": 3.408392913454423e-05, - "loss": 2.3915, + "epoch": 0.5381122362254438, + "grad_norm": 0.73828125, + "learning_rate": 0.00010413784965660063, + "loss": 2.4134, "step": 1385 }, { - "epoch": 0.44387510008006403, - "grad_norm": 0.412109375, - "learning_rate": 3.4057874468573145e-05, - "loss": 1.9231, + "epoch": 0.5385007649158593, + "grad_norm": 0.6953125, + "learning_rate": 0.00010400225616817895, + "loss": 2.3331, "step": 1386 }, { - "epoch": 0.44419535628502804, - "grad_norm": 0.546875, - "learning_rate": 3.4031808472794875e-05, - "loss": 2.1587, + "epoch": 0.5388892936062747, + "grad_norm": 0.75, + "learning_rate": 0.00010386665530916191, + "loss": 2.4752, "step": 1387 }, { - "epoch": 0.444515612489992, - "grad_norm": 0.53515625, - "learning_rate": 3.400573117981338e-05, - "loss": 1.8927, + "epoch": 0.5392778222966902, + "grad_norm": 0.69140625, + "learning_rate": 0.00010373104732927341, + "loss": 2.396, "step": 1388 }, { - "epoch": 0.44483586869495595, - "grad_norm": 0.46484375, - "learning_rate": 3.3979642622246765e-05, - "loss": 1.9999, + "epoch": 0.5396663509871057, + "grad_norm": 0.71484375, + "learning_rate": 0.00010359543247825045, + "loss": 2.3771, "step": 1389 }, { - "epoch": 0.44515612489991996, - "grad_norm": 0.515625, - "learning_rate": 3.395354283272724e-05, - "loss": 1.7148, + "epoch": 0.5400548796775212, + "grad_norm": 0.69921875, + "learning_rate": 0.00010345981100584272, + "loss": 2.3218, "step": 1390 }, { - "epoch": 0.4454763811048839, - "grad_norm": 0.54296875, - "learning_rate": 3.3927431843901034e-05, - "loss": 2.6938, + "epoch": 0.5404434083679367, + "grad_norm": 0.67578125, + "learning_rate": 0.0001033241831618121, + "loss": 2.2933, "step": 1391 }, { - "epoch": 0.44579663730984787, - "grad_norm": 0.58203125, - "learning_rate": 3.39013096884284e-05, - "loss": 2.2425, + "epoch": 0.5408319370583522, + "grad_norm": 0.71875, + "learning_rate": 0.00010318854919593221, + "loss": 2.4244, "step": 1392 }, { - "epoch": 0.4461168935148119, - "grad_norm": 0.431640625, - "learning_rate": 3.3875176398983565e-05, - "loss": 2.3499, + "epoch": 0.5412204657487676, + "grad_norm": 0.67578125, + "learning_rate": 0.00010305290935798792, + "loss": 2.3525, "step": 1393 }, { - "epoch": 0.4464371497197758, - "grad_norm": 0.455078125, - "learning_rate": 3.3849032008254676e-05, - "loss": 2.2984, + "epoch": 0.5416089944391831, + "grad_norm": 0.68359375, + "learning_rate": 0.00010291726389777489, + "loss": 2.3758, "step": 1394 }, { - "epoch": 0.4467574059247398, - "grad_norm": 6.21875, - "learning_rate": 3.382287654894377e-05, - "loss": 1.9116, + "epoch": 0.5419975231295986, + "grad_norm": 0.73828125, + "learning_rate": 0.00010278161306509923, + "loss": 2.3396, "step": 1395 }, { - "epoch": 0.4470776621297038, - "grad_norm": 0.67578125, - "learning_rate": 3.379671005376671e-05, - "loss": 2.2836, + "epoch": 0.5423860518200141, + "grad_norm": 0.7265625, + "learning_rate": 0.00010264595710977687, + "loss": 2.3817, "step": 1396 }, { - "epoch": 0.44739791833466774, - "grad_norm": 0.328125, - "learning_rate": 3.377053255545319e-05, - "loss": 1.6366, + "epoch": 0.5427745805104296, + "grad_norm": 0.6796875, + "learning_rate": 0.00010251029628163316, + "loss": 2.4236, "step": 1397 }, { - "epoch": 0.4477181745396317, - "grad_norm": 0.470703125, - "learning_rate": 3.374434408674665e-05, - "loss": 2.4968, + "epoch": 0.5431631092008451, + "grad_norm": 0.6953125, + "learning_rate": 0.00010237463083050245, + "loss": 2.3765, "step": 1398 }, { - "epoch": 0.44803843074459565, - "grad_norm": 0.37109375, - "learning_rate": 3.371814468040426e-05, - "loss": 2.4331, + "epoch": 0.5435516378912605, + "grad_norm": 0.71875, + "learning_rate": 0.0001022389610062276, + "loss": 2.3717, "step": 1399 }, { - "epoch": 0.44835868694955966, - "grad_norm": 0.50390625, - "learning_rate": 3.3691934369196855e-05, - "loss": 2.381, + "epoch": 0.543940166581676, + "grad_norm": 0.73046875, + "learning_rate": 0.00010210328705865962, + "loss": 2.3726, "step": 1400 }, { - "epoch": 0.4486789431545236, - "grad_norm": 0.87109375, - "learning_rate": 3.366571318590894e-05, - "loss": 1.7545, + "epoch": 0.5443286952720915, + "grad_norm": 0.6640625, + "learning_rate": 0.00010196760923765689, + "loss": 2.3417, "step": 1401 }, { - "epoch": 0.44899919935948757, - "grad_norm": 0.380859375, - "learning_rate": 3.3639481163338584e-05, - "loss": 2.4128, + "epoch": 0.544717223962507, + "grad_norm": 0.7421875, + "learning_rate": 0.00010183192779308512, + "loss": 2.3906, "step": 1402 }, { - "epoch": 0.4493194555644516, - "grad_norm": 0.37890625, - "learning_rate": 3.3613238334297446e-05, - "loss": 1.8779, + "epoch": 0.5451057526529225, + "grad_norm": 0.71875, + "learning_rate": 0.00010169624297481664, + "loss": 2.3572, "step": 1403 }, { - "epoch": 0.44963971176941553, - "grad_norm": 0.447265625, - "learning_rate": 3.358698473161068e-05, - "loss": 2.3488, + "epoch": 0.545494281343338, + "grad_norm": 0.6953125, + "learning_rate": 0.00010156055503272996, + "loss": 2.3263, "step": 1404 }, { - "epoch": 0.4499599679743795, - "grad_norm": 0.54296875, - "learning_rate": 3.3560720388116934e-05, - "loss": 1.6194, + "epoch": 0.5458828100337534, + "grad_norm": 0.76171875, + "learning_rate": 0.00010142486421670941, + "loss": 2.3674, "step": 1405 }, { - "epoch": 0.4502802241793435, - "grad_norm": 1.2421875, - "learning_rate": 3.3534445336668266e-05, - "loss": 2.0731, + "epoch": 0.5462713387241689, + "grad_norm": 0.6875, + "learning_rate": 0.00010128917077664448, + "loss": 2.3493, "step": 1406 }, { - "epoch": 0.45060048038430744, - "grad_norm": 0.349609375, - "learning_rate": 3.350815961013014e-05, - "loss": 2.0653, + "epoch": 0.5466598674145844, + "grad_norm": 0.76171875, + "learning_rate": 0.00010115347496242964, + "loss": 2.4162, "step": 1407 }, { - "epoch": 0.4509207365892714, - "grad_norm": 0.56640625, - "learning_rate": 3.348186324138139e-05, - "loss": 2.5686, + "epoch": 0.5470483961049999, + "grad_norm": 0.73828125, + "learning_rate": 0.00010101777702396367, + "loss": 2.4186, "step": 1408 }, { - "epoch": 0.4512409927942354, - "grad_norm": 0.59375, - "learning_rate": 3.345555626331415e-05, - "loss": 2.0038, + "epoch": 0.5474369247954154, + "grad_norm": 0.66796875, + "learning_rate": 0.00010088207721114922, + "loss": 2.365, "step": 1409 }, { - "epoch": 0.45156124899919936, - "grad_norm": 0.56640625, - "learning_rate": 3.34292387088338e-05, - "loss": 2.3573, + "epoch": 0.5478254534858309, + "grad_norm": 0.69921875, + "learning_rate": 0.00010074637577389253, + "loss": 2.3712, "step": 1410 }, { - "epoch": 0.4518815052041633, - "grad_norm": 0.56640625, - "learning_rate": 3.340291061085898e-05, - "loss": 2.2184, + "epoch": 0.5482139821762463, + "grad_norm": 0.67578125, + "learning_rate": 0.0001006106729621026, + "loss": 2.3686, "step": 1411 }, { - "epoch": 0.4522017614091273, - "grad_norm": 0.5703125, - "learning_rate": 3.3376572002321506e-05, - "loss": 2.8585, + "epoch": 0.5486025108666618, + "grad_norm": 0.6953125, + "learning_rate": 0.00010047496902569122, + "loss": 2.3628, "step": 1412 }, { - "epoch": 0.4525220176140913, - "grad_norm": 0.400390625, - "learning_rate": 3.335022291616635e-05, - "loss": 2.5615, + "epoch": 0.5489910395570773, + "grad_norm": 0.66796875, + "learning_rate": 0.00010033926421457207, + "loss": 2.34, "step": 1413 }, { - "epoch": 0.45284227381905523, - "grad_norm": 0.734375, - "learning_rate": 3.332386338535158e-05, - "loss": 2.1242, + "epoch": 0.5493795682474928, + "grad_norm": 0.76171875, + "learning_rate": 0.00010020355877866054, + "loss": 2.359, "step": 1414 }, { - "epoch": 0.45316253002401924, - "grad_norm": 0.88671875, - "learning_rate": 3.329749344284832e-05, - "loss": 2.2279, + "epoch": 0.5497680969379083, + "grad_norm": 0.69140625, + "learning_rate": 0.0001000678529678731, + "loss": 2.3546, "step": 1415 }, { - "epoch": 0.4534827862289832, - "grad_norm": 0.64453125, - "learning_rate": 3.327111312164075e-05, - "loss": 2.3805, + "epoch": 0.5501566256283238, + "grad_norm": 0.69140625, + "learning_rate": 9.993214703212696e-05, + "loss": 2.3523, "step": 1416 }, { - "epoch": 0.45380304243394715, - "grad_norm": 0.41796875, - "learning_rate": 3.324472245472598e-05, - "loss": 2.2084, + "epoch": 0.5505451543187392, + "grad_norm": 0.72265625, + "learning_rate": 9.97964412213395e-05, + "loss": 2.3794, "step": 1417 }, { - "epoch": 0.45412329863891115, - "grad_norm": 0.5390625, - "learning_rate": 3.321832147511412e-05, - "loss": 2.6584, + "epoch": 0.5509336830091547, + "grad_norm": 0.83984375, + "learning_rate": 9.966073578542795e-05, + "loss": 2.3968, "step": 1418 }, { - "epoch": 0.4544435548438751, - "grad_norm": 0.5078125, - "learning_rate": 3.319191021582814e-05, - "loss": 1.9955, + "epoch": 0.5513222116995702, + "grad_norm": 0.7421875, + "learning_rate": 9.952503097430882e-05, + "loss": 2.4061, "step": 1419 }, { - "epoch": 0.45476381104883906, - "grad_norm": 0.455078125, - "learning_rate": 3.3165488709903876e-05, - "loss": 2.1134, + "epoch": 0.5517107403899857, + "grad_norm": 0.73046875, + "learning_rate": 9.938932703789743e-05, + "loss": 2.3675, "step": 1420 }, { - "epoch": 0.455084067253803, - "grad_norm": 0.474609375, - "learning_rate": 3.313905699038997e-05, - "loss": 2.4377, + "epoch": 0.5520992690804012, + "grad_norm": 0.6953125, + "learning_rate": 9.925362422610754e-05, + "loss": 2.3192, "step": 1421 }, { - "epoch": 0.455404323458767, - "grad_norm": 0.498046875, - "learning_rate": 3.3112615090347875e-05, - "loss": 2.3482, + "epoch": 0.5524877977708167, + "grad_norm": 0.71484375, + "learning_rate": 9.91179227888508e-05, + "loss": 2.354, "step": 1422 }, { - "epoch": 0.455724579663731, - "grad_norm": 0.416015625, - "learning_rate": 3.3086163042851757e-05, - "loss": 2.2396, + "epoch": 0.5528763264612321, + "grad_norm": 0.67578125, + "learning_rate": 9.898222297603639e-05, + "loss": 2.3639, "step": 1423 }, { - "epoch": 0.45604483586869493, - "grad_norm": 0.63671875, - "learning_rate": 3.305970088098844e-05, - "loss": 1.9499, + "epoch": 0.5532648551516476, + "grad_norm": 0.7109375, + "learning_rate": 9.884652503757041e-05, + "loss": 2.4172, "step": 1424 }, { - "epoch": 0.45636509207365894, - "grad_norm": 0.40234375, - "learning_rate": 3.303322863785747e-05, - "loss": 2.4442, + "epoch": 0.5536533838420631, + "grad_norm": 0.69140625, + "learning_rate": 9.871082922335554e-05, + "loss": 2.3958, "step": 1425 }, { - "epoch": 0.4566853482786229, - "grad_norm": 0.5078125, - "learning_rate": 3.300674634657094e-05, - "loss": 2.0177, + "epoch": 0.5540419125324786, + "grad_norm": 0.6953125, + "learning_rate": 9.857513578329063e-05, + "loss": 2.3441, "step": 1426 }, { - "epoch": 0.45700560448358685, - "grad_norm": 0.3203125, - "learning_rate": 3.2980254040253554e-05, - "loss": 2.3229, + "epoch": 0.5544304412228941, + "grad_norm": 0.65625, + "learning_rate": 9.843944496727007e-05, + "loss": 2.3684, "step": 1427 }, { - "epoch": 0.45732586068855086, - "grad_norm": 0.478515625, - "learning_rate": 3.2953751752042514e-05, - "loss": 1.5711, + "epoch": 0.5548189699133096, + "grad_norm": 0.76171875, + "learning_rate": 9.830375702518338e-05, + "loss": 2.3754, "step": 1428 }, { - "epoch": 0.4576461168935148, - "grad_norm": 0.486328125, - "learning_rate": 3.2927239515087524e-05, - "loss": 2.3031, + "epoch": 0.555207498603725, + "grad_norm": 0.67578125, + "learning_rate": 9.81680722069149e-05, + "loss": 2.3649, "step": 1429 }, { - "epoch": 0.45796637309847876, - "grad_norm": 1.03125, - "learning_rate": 3.290071736255073e-05, - "loss": 2.4136, + "epoch": 0.5555960272941405, + "grad_norm": 0.69140625, + "learning_rate": 9.803239076234312e-05, + "loss": 2.4059, "step": 1430 }, { - "epoch": 0.45828662930344277, - "grad_norm": 0.412109375, - "learning_rate": 3.287418532760667e-05, - "loss": 2.4615, + "epoch": 0.555984555984556, + "grad_norm": 0.7109375, + "learning_rate": 9.78967129413404e-05, + "loss": 2.3236, "step": 1431 }, { - "epoch": 0.4586068855084067, - "grad_norm": 0.5078125, - "learning_rate": 3.284764344344226e-05, - "loss": 2.3343, + "epoch": 0.5563730846749715, + "grad_norm": 0.6953125, + "learning_rate": 9.77610389937724e-05, + "loss": 2.3588, "step": 1432 }, { - "epoch": 0.4589271417133707, - "grad_norm": 0.76171875, - "learning_rate": 3.282109174325672e-05, - "loss": 2.6001, + "epoch": 0.556761613365387, + "grad_norm": 0.74609375, + "learning_rate": 9.762536916949759e-05, + "loss": 2.4225, "step": 1433 }, { - "epoch": 0.4592473979183347, - "grad_norm": 0.75, - "learning_rate": 3.279453026026157e-05, - "loss": 2.282, + "epoch": 0.5571501420558025, + "grad_norm": 0.73046875, + "learning_rate": 9.748970371836688e-05, + "loss": 2.4599, "step": 1434 }, { - "epoch": 0.45956765412329864, - "grad_norm": 0.68359375, - "learning_rate": 3.2767959027680525e-05, - "loss": 2.6124, + "epoch": 0.5575386707462179, + "grad_norm": 0.66796875, + "learning_rate": 9.735404289022316e-05, + "loss": 2.2971, "step": 1435 }, { - "epoch": 0.4598879103282626, - "grad_norm": 0.47265625, - "learning_rate": 3.2741378078749545e-05, - "loss": 1.9807, + "epoch": 0.5579271994366334, + "grad_norm": 0.6796875, + "learning_rate": 9.721838693490079e-05, + "loss": 2.2642, "step": 1436 }, { - "epoch": 0.4602081665332266, - "grad_norm": 0.6328125, - "learning_rate": 3.271478744671672e-05, - "loss": 2.7701, + "epoch": 0.5583157281270489, + "grad_norm": 0.671875, + "learning_rate": 9.708273610222512e-05, + "loss": 2.3131, "step": 1437 }, { - "epoch": 0.46052842273819056, - "grad_norm": 2.09375, - "learning_rate": 3.2688187164842235e-05, - "loss": 2.2759, + "epoch": 0.5587042568174644, + "grad_norm": 0.74609375, + "learning_rate": 9.69470906420121e-05, + "loss": 2.3757, "step": 1438 }, { - "epoch": 0.4608486789431545, - "grad_norm": 0.474609375, - "learning_rate": 3.266157726639836e-05, - "loss": 2.1833, + "epoch": 0.5590927855078799, + "grad_norm": 0.69921875, + "learning_rate": 9.68114508040678e-05, + "loss": 2.3271, "step": 1439 }, { - "epoch": 0.4611689351481185, - "grad_norm": 0.64453125, - "learning_rate": 3.263495778466942e-05, - "loss": 2.2356, + "epoch": 0.5594813141982953, + "grad_norm": 0.6953125, + "learning_rate": 9.667581683818791e-05, + "loss": 2.3897, "step": 1440 }, { - "epoch": 0.4614891913530825, - "grad_norm": 0.94140625, - "learning_rate": 3.260832875295169e-05, - "loss": 2.2155, + "epoch": 0.5598698428887108, + "grad_norm": 0.703125, + "learning_rate": 9.65401889941573e-05, + "loss": 2.4096, "step": 1441 }, { - "epoch": 0.4618094475580464, - "grad_norm": 0.349609375, - "learning_rate": 3.25816902045534e-05, - "loss": 1.4778, + "epoch": 0.5602583715791263, + "grad_norm": 0.67578125, + "learning_rate": 9.640456752174957e-05, + "loss": 2.3779, "step": 1442 }, { - "epoch": 0.46212970376301044, - "grad_norm": 0.447265625, - "learning_rate": 3.255504217279469e-05, - "loss": 1.7559, + "epoch": 0.5606469002695418, + "grad_norm": 0.69140625, + "learning_rate": 9.626895267072662e-05, + "loss": 2.3463, "step": 1443 }, { - "epoch": 0.4624499599679744, - "grad_norm": 0.515625, - "learning_rate": 3.2528384691007555e-05, - "loss": 2.1804, + "epoch": 0.5610354289599573, + "grad_norm": 0.67578125, + "learning_rate": 9.613334469083811e-05, + "loss": 2.2979, "step": 1444 }, { - "epoch": 0.46277021617293834, - "grad_norm": 0.69140625, - "learning_rate": 3.250171779253583e-05, - "loss": 2.2981, + "epoch": 0.5614239576503728, + "grad_norm": 0.65625, + "learning_rate": 9.599774383182106e-05, + "loss": 2.3747, "step": 1445 }, { - "epoch": 0.4630904723779023, - "grad_norm": 0.48046875, - "learning_rate": 3.2475041510735086e-05, - "loss": 2.4689, + "epoch": 0.5618124863407882, + "grad_norm": 0.7578125, + "learning_rate": 9.586215034339939e-05, + "loss": 2.3214, "step": 1446 }, { - "epoch": 0.4634107285828663, - "grad_norm": 0.43359375, - "learning_rate": 3.244835587897268e-05, - "loss": 2.397, + "epoch": 0.5622010150312037, + "grad_norm": 0.7578125, + "learning_rate": 9.572656447528345e-05, + "loss": 2.398, "step": 1447 }, { - "epoch": 0.46373098478783026, - "grad_norm": 0.482421875, - "learning_rate": 3.242166093062764e-05, - "loss": 2.2169, + "epoch": 0.5625895437216192, + "grad_norm": 0.765625, + "learning_rate": 9.559098647716952e-05, + "loss": 2.4058, "step": 1448 }, { - "epoch": 0.4640512409927942, - "grad_norm": 0.57421875, - "learning_rate": 3.239495669909064e-05, - "loss": 2.2407, + "epoch": 0.5629780724120347, + "grad_norm": 0.72265625, + "learning_rate": 9.545541659873945e-05, + "loss": 2.425, "step": 1449 }, { - "epoch": 0.4643714971977582, - "grad_norm": 0.57421875, - "learning_rate": 3.236824321776397e-05, - "loss": 2.1264, + "epoch": 0.5633666011024502, + "grad_norm": 0.71875, + "learning_rate": 9.531985508966004e-05, + "loss": 2.3954, "step": 1450 }, { - "epoch": 0.4646917534027222, - "grad_norm": 0.46875, - "learning_rate": 3.2341520520061516e-05, - "loss": 2.2125, + "epoch": 0.5637551297928657, + "grad_norm": 0.78515625, + "learning_rate": 9.518430219958277e-05, + "loss": 2.3367, "step": 1451 }, { - "epoch": 0.4650120096076861, - "grad_norm": 0.435546875, - "learning_rate": 3.2314788639408656e-05, - "loss": 2.1495, + "epoch": 0.5641436584832811, + "grad_norm": 0.71484375, + "learning_rate": 9.504875817814321e-05, + "loss": 2.3757, "step": 1452 }, { - "epoch": 0.46533226581265014, - "grad_norm": 0.466796875, - "learning_rate": 3.2288047609242266e-05, - "loss": 2.029, + "epoch": 0.5645321871736966, + "grad_norm": 0.66796875, + "learning_rate": 9.491322327496062e-05, + "loss": 2.3418, "step": 1453 }, { - "epoch": 0.4656525220176141, - "grad_norm": 0.625, - "learning_rate": 3.226129746301067e-05, - "loss": 1.9174, + "epoch": 0.5649207158641121, + "grad_norm": 0.6875, + "learning_rate": 9.477769773963738e-05, + "loss": 2.3674, "step": 1454 }, { - "epoch": 0.46597277822257804, - "grad_norm": 0.494140625, - "learning_rate": 3.223453823417362e-05, - "loss": 2.5737, + "epoch": 0.5653092445545276, + "grad_norm": 0.73828125, + "learning_rate": 9.464218182175875e-05, + "loss": 2.3464, "step": 1455 }, { - "epoch": 0.46629303442754205, - "grad_norm": 0.62109375, - "learning_rate": 3.220776995620217e-05, - "loss": 2.2023, + "epoch": 0.5656977732449431, + "grad_norm": 0.72265625, + "learning_rate": 9.450667577089217e-05, + "loss": 2.3947, "step": 1456 }, { - "epoch": 0.466613290632506, - "grad_norm": 0.4375, - "learning_rate": 3.218099266257874e-05, - "loss": 2.6245, + "epoch": 0.5660863019353586, + "grad_norm": 0.70703125, + "learning_rate": 9.437117983658701e-05, + "loss": 2.3727, "step": 1457 }, { - "epoch": 0.46693354683746996, - "grad_norm": 0.455078125, - "learning_rate": 3.2154206386797e-05, - "loss": 1.9389, + "epoch": 0.566474830625774, + "grad_norm": 0.6875, + "learning_rate": 9.423569426837388e-05, + "loss": 2.4071, "step": 1458 }, { - "epoch": 0.46725380304243397, - "grad_norm": 0.53125, - "learning_rate": 3.212741116236189e-05, - "loss": 2.0921, + "epoch": 0.5668633593161895, + "grad_norm": 0.71484375, + "learning_rate": 9.41002193157644e-05, + "loss": 2.3527, "step": 1459 }, { - "epoch": 0.4675740592473979, - "grad_norm": 0.5078125, - "learning_rate": 3.210060702278951e-05, - "loss": 2.5264, + "epoch": 0.567251888006605, + "grad_norm": 0.7578125, + "learning_rate": 9.396475522825066e-05, + "loss": 2.3901, "step": 1460 }, { - "epoch": 0.4678943154523619, - "grad_norm": 0.87890625, - "learning_rate": 3.207379400160711e-05, - "loss": 1.9612, + "epoch": 0.5676404166970205, + "grad_norm": 0.734375, + "learning_rate": 9.382930225530462e-05, + "loss": 2.4118, "step": 1461 }, { - "epoch": 0.4682145716573259, - "grad_norm": 0.482421875, - "learning_rate": 3.204697213235309e-05, - "loss": 2.329, + "epoch": 0.568028945387436, + "grad_norm": 0.86328125, + "learning_rate": 9.369386064637785e-05, + "loss": 2.4239, "step": 1462 }, { - "epoch": 0.46853482786228984, - "grad_norm": 0.58203125, - "learning_rate": 3.2020141448576866e-05, - "loss": 2.4217, + "epoch": 0.5684174740778515, + "grad_norm": 0.6796875, + "learning_rate": 9.355843065090104e-05, + "loss": 2.4176, "step": 1463 }, { - "epoch": 0.4688550840672538, - "grad_norm": 0.453125, - "learning_rate": 3.1993301983838906e-05, - "loss": 1.7458, + "epoch": 0.5688060027682669, + "grad_norm": 0.78515625, + "learning_rate": 9.342301251828336e-05, + "loss": 2.4052, "step": 1464 }, { - "epoch": 0.4691753402722178, - "grad_norm": 0.3828125, - "learning_rate": 3.1966453771710655e-05, - "loss": 2.112, + "epoch": 0.5691945314586824, + "grad_norm": 0.7109375, + "learning_rate": 9.328760649791227e-05, + "loss": 2.4543, "step": 1465 }, { - "epoch": 0.46949559647718175, - "grad_norm": 0.546875, - "learning_rate": 3.1939596845774525e-05, - "loss": 2.4703, + "epoch": 0.5695830601490979, + "grad_norm": 0.81640625, + "learning_rate": 9.315221283915282e-05, + "loss": 2.3569, "step": 1466 }, { - "epoch": 0.4698158526821457, - "grad_norm": 0.71875, - "learning_rate": 3.191273123962378e-05, - "loss": 2.4275, + "epoch": 0.5699715888395134, + "grad_norm": 0.7109375, + "learning_rate": 9.301683179134741e-05, + "loss": 2.4172, "step": 1467 }, { - "epoch": 0.47013610888710966, - "grad_norm": 0.53125, - "learning_rate": 3.188585698686257e-05, - "loss": 2.4174, + "epoch": 0.5703601175299289, + "grad_norm": 0.76953125, + "learning_rate": 9.288146360381507e-05, + "loss": 2.4128, "step": 1468 }, { - "epoch": 0.47045636509207367, - "grad_norm": 0.6328125, - "learning_rate": 3.185897412110587e-05, - "loss": 2.2741, + "epoch": 0.5707486462203444, + "grad_norm": 0.765625, + "learning_rate": 9.274610852585127e-05, + "loss": 2.347, "step": 1469 }, { - "epoch": 0.4707766212970376, - "grad_norm": 0.80078125, - "learning_rate": 3.1832082675979394e-05, - "loss": 2.2359, + "epoch": 0.5711371749107598, + "grad_norm": 0.74609375, + "learning_rate": 9.261076680672724e-05, + "loss": 2.3291, "step": 1470 }, { - "epoch": 0.4710968775020016, - "grad_norm": 0.6953125, - "learning_rate": 3.180518268511963e-05, - "loss": 2.174, + "epoch": 0.5715257036011753, + "grad_norm": 0.80859375, + "learning_rate": 9.24754386956897e-05, + "loss": 2.3808, "step": 1471 }, { - "epoch": 0.4714171337069656, - "grad_norm": 0.447265625, - "learning_rate": 3.1778274182173716e-05, - "loss": 2.2126, + "epoch": 0.5719142322915908, + "grad_norm": 0.69921875, + "learning_rate": 9.234012444196029e-05, + "loss": 2.3653, "step": 1472 }, { - "epoch": 0.47173738991192954, - "grad_norm": 0.392578125, - "learning_rate": 3.175135720079947e-05, - "loss": 2.0206, + "epoch": 0.5723027609820063, + "grad_norm": 0.80078125, + "learning_rate": 9.220482429473502e-05, + "loss": 2.3972, "step": 1473 }, { - "epoch": 0.4720576461168935, - "grad_norm": 0.74609375, - "learning_rate": 3.172443177466529e-05, - "loss": 2.4025, + "epoch": 0.5726912896724218, + "grad_norm": 0.6640625, + "learning_rate": 9.206953850318408e-05, + "loss": 2.3442, "step": 1474 }, { - "epoch": 0.4723779023218575, - "grad_norm": 0.73828125, - "learning_rate": 3.169749793745014e-05, - "loss": 2.6801, + "epoch": 0.5730798183628373, + "grad_norm": 0.74609375, + "learning_rate": 9.193426731645112e-05, + "loss": 2.4275, "step": 1475 }, { - "epoch": 0.47269815852682145, - "grad_norm": 0.55859375, - "learning_rate": 3.167055572284352e-05, - "loss": 2.2362, + "epoch": 0.5734683470532527, + "grad_norm": 0.71875, + "learning_rate": 9.179901098365297e-05, + "loss": 2.3666, "step": 1476 }, { - "epoch": 0.4730184147317854, - "grad_norm": 0.43359375, - "learning_rate": 3.164360516454541e-05, - "loss": 2.0306, + "epoch": 0.5738568757436682, + "grad_norm": 0.71484375, + "learning_rate": 9.166376975387898e-05, + "loss": 2.3713, "step": 1477 }, { - "epoch": 0.4733386709367494, - "grad_norm": 0.431640625, - "learning_rate": 3.1616646296266196e-05, - "loss": 2.1496, + "epoch": 0.5742454044340837, + "grad_norm": 0.74609375, + "learning_rate": 9.152854387619081e-05, + "loss": 2.399, "step": 1478 }, { - "epoch": 0.47365892714171337, - "grad_norm": 1.171875, - "learning_rate": 3.158967915172669e-05, - "loss": 2.6622, + "epoch": 0.5746339331244992, + "grad_norm": 0.734375, + "learning_rate": 9.139333359962179e-05, + "loss": 2.35, "step": 1479 }, { - "epoch": 0.4739791833466773, - "grad_norm": 0.80078125, - "learning_rate": 3.1562703764658055e-05, - "loss": 2.2467, + "epoch": 0.5750224618149147, + "grad_norm": 0.6484375, + "learning_rate": 9.125813917317656e-05, + "loss": 2.3904, "step": 1480 }, { - "epoch": 0.47429943955164133, - "grad_norm": 0.349609375, - "learning_rate": 3.1535720168801725e-05, - "loss": 1.7394, + "epoch": 0.5754109905053302, + "grad_norm": 0.7421875, + "learning_rate": 9.112296084583051e-05, + "loss": 2.4128, "step": 1481 }, { - "epoch": 0.4746196957566053, - "grad_norm": 0.462890625, - "learning_rate": 3.150872839790946e-05, - "loss": 1.9964, + "epoch": 0.5757995191957456, + "grad_norm": 0.68359375, + "learning_rate": 9.098779886652941e-05, + "loss": 2.3246, "step": 1482 }, { - "epoch": 0.47493995196156924, - "grad_norm": 0.5703125, - "learning_rate": 3.148172848574319e-05, - "loss": 2.3196, + "epoch": 0.5761880478861611, + "grad_norm": 0.6875, + "learning_rate": 9.085265348418894e-05, + "loss": 2.3642, "step": 1483 }, { - "epoch": 0.47526020816653325, - "grad_norm": 0.40234375, - "learning_rate": 3.145472046607508e-05, - "loss": 2.052, + "epoch": 0.5765765765765766, + "grad_norm": 0.67578125, + "learning_rate": 9.071752494769422e-05, + "loss": 2.3825, "step": 1484 }, { - "epoch": 0.4755804643714972, - "grad_norm": 0.4296875, - "learning_rate": 3.142770437268739e-05, - "loss": 1.8644, + "epoch": 0.5769651052669921, + "grad_norm": 0.6953125, + "learning_rate": 9.058241350589929e-05, + "loss": 2.3952, "step": 1485 }, { - "epoch": 0.47590072057646116, - "grad_norm": 0.83203125, - "learning_rate": 3.140068023937251e-05, - "loss": 2.1707, + "epoch": 0.5773536339574076, + "grad_norm": 0.69140625, + "learning_rate": 9.044731940762675e-05, + "loss": 2.3592, "step": 1486 }, { - "epoch": 0.47622097678142516, - "grad_norm": 0.59375, - "learning_rate": 3.137364809993288e-05, - "loss": 1.9486, + "epoch": 0.5777421626478231, + "grad_norm": 0.69921875, + "learning_rate": 9.031224290166727e-05, + "loss": 2.4437, "step": 1487 }, { - "epoch": 0.4765412329863891, - "grad_norm": 0.431640625, - "learning_rate": 3.1346607988180933e-05, - "loss": 1.5815, + "epoch": 0.5781306913382385, + "grad_norm": 0.7109375, + "learning_rate": 9.017718423677911e-05, + "loss": 2.4009, "step": 1488 }, { - "epoch": 0.47686148919135307, - "grad_norm": 0.59375, - "learning_rate": 3.131955993793911e-05, - "loss": 2.0043, + "epoch": 0.578519220028654, + "grad_norm": 0.6796875, + "learning_rate": 9.004214366168769e-05, + "loss": 2.3155, "step": 1489 }, { - "epoch": 0.4771817453963171, - "grad_norm": 0.546875, - "learning_rate": 3.129250398303975e-05, - "loss": 2.2674, + "epoch": 0.5789077487190695, + "grad_norm": 0.70703125, + "learning_rate": 8.990712142508505e-05, + "loss": 2.4467, "step": 1490 }, { - "epoch": 0.47750200160128103, - "grad_norm": 0.3828125, - "learning_rate": 3.12654401573251e-05, - "loss": 2.037, + "epoch": 0.579296277409485, + "grad_norm": 0.734375, + "learning_rate": 8.977211777562953e-05, + "loss": 2.3164, "step": 1491 }, { - "epoch": 0.477822257806245, - "grad_norm": 0.427734375, - "learning_rate": 3.123836849464723e-05, - "loss": 1.7428, + "epoch": 0.5796848060999005, + "grad_norm": 0.703125, + "learning_rate": 8.963713296194521e-05, + "loss": 2.3961, "step": 1492 }, { - "epoch": 0.47814251401120894, - "grad_norm": 0.921875, - "learning_rate": 3.121128902886804e-05, - "loss": 2.6863, + "epoch": 0.580073334790316, + "grad_norm": 0.68359375, + "learning_rate": 8.950216723262152e-05, + "loss": 2.3545, "step": 1493 }, { - "epoch": 0.47846277021617295, - "grad_norm": 0.515625, - "learning_rate": 3.118420179385919e-05, - "loss": 2.0104, + "epoch": 0.5804618634807314, + "grad_norm": 0.67578125, + "learning_rate": 8.936722083621265e-05, + "loss": 2.4041, "step": 1494 }, { - "epoch": 0.4787830264211369, - "grad_norm": 0.5234375, - "learning_rate": 3.115710682350201e-05, - "loss": 2.1841, + "epoch": 0.5808503921711469, + "grad_norm": 0.6875, + "learning_rate": 8.923229402123728e-05, + "loss": 2.3851, "step": 1495 }, { - "epoch": 0.47910328262610086, - "grad_norm": 0.61328125, - "learning_rate": 3.113000415168756e-05, - "loss": 2.503, + "epoch": 0.5812389208615624, + "grad_norm": 0.74609375, + "learning_rate": 8.909738703617799e-05, + "loss": 2.3893, "step": 1496 }, { - "epoch": 0.47942353883106487, - "grad_norm": 0.32421875, - "learning_rate": 3.110289381231651e-05, - "loss": 1.9073, + "epoch": 0.5816274495519779, + "grad_norm": 0.765625, + "learning_rate": 8.896250012948086e-05, + "loss": 2.2755, "step": 1497 }, { - "epoch": 0.4797437950360288, - "grad_norm": 1.421875, - "learning_rate": 3.107577583929914e-05, - "loss": 1.9962, + "epoch": 0.5820159782423934, + "grad_norm": 0.70703125, + "learning_rate": 8.882763354955495e-05, + "loss": 2.4155, "step": 1498 }, { - "epoch": 0.4800640512409928, - "grad_norm": 0.48046875, - "learning_rate": 3.104865026655525e-05, - "loss": 2.0423, + "epoch": 0.5824045069328089, + "grad_norm": 0.66015625, + "learning_rate": 8.869278754477193e-05, + "loss": 2.382, "step": 1499 }, { - "epoch": 0.4803843074459568, - "grad_norm": 0.41796875, - "learning_rate": 3.102151712801415e-05, - "loss": 2.1058, + "epoch": 0.5827930356232243, + "grad_norm": 0.65625, + "learning_rate": 8.855796236346557e-05, + "loss": 2.3534, "step": 1500 }, { - "epoch": 0.48070456365092074, - "grad_norm": 0.60546875, - "learning_rate": 3.0994376457614636e-05, - "loss": 2.2794, + "epoch": 0.5831815643136398, + "grad_norm": 0.65234375, + "learning_rate": 8.84231582539313e-05, + "loss": 2.3147, "step": 1501 }, { - "epoch": 0.4810248198558847, - "grad_norm": 0.98046875, - "learning_rate": 3.09672282893049e-05, - "loss": 2.1654, + "epoch": 0.5835700930040553, + "grad_norm": 0.72265625, + "learning_rate": 8.82883754644257e-05, + "loss": 2.3962, "step": 1502 }, { - "epoch": 0.4813450760608487, - "grad_norm": 0.40625, - "learning_rate": 3.094007265704253e-05, - "loss": 1.8414, + "epoch": 0.5839586216944708, + "grad_norm": 0.6640625, + "learning_rate": 8.815361424316616e-05, + "loss": 2.3999, "step": 1503 }, { - "epoch": 0.48166533226581265, - "grad_norm": 0.470703125, - "learning_rate": 3.091290959479444e-05, - "loss": 1.9365, + "epoch": 0.5843471503848863, + "grad_norm": 0.70703125, + "learning_rate": 8.801887483833023e-05, + "loss": 2.3827, "step": 1504 }, { - "epoch": 0.4819855884707766, - "grad_norm": 1.1796875, - "learning_rate": 3.0885739136536854e-05, - "loss": 2.2331, + "epoch": 0.5847356790753017, + "grad_norm": 0.6640625, + "learning_rate": 8.788415749805541e-05, + "loss": 2.3867, "step": 1505 }, { - "epoch": 0.4823058446757406, - "grad_norm": 0.435546875, - "learning_rate": 3.0858561316255224e-05, - "loss": 1.8893, + "epoch": 0.5851242077657172, + "grad_norm": 0.6796875, + "learning_rate": 8.774946247043852e-05, + "loss": 2.3742, "step": 1506 }, { - "epoch": 0.48262610088070457, - "grad_norm": 0.34375, - "learning_rate": 3.083137616794422e-05, - "loss": 1.8019, + "epoch": 0.5855127364561327, + "grad_norm": 0.6953125, + "learning_rate": 8.761479000353528e-05, + "loss": 2.3596, "step": 1507 }, { - "epoch": 0.4829463570856685, - "grad_norm": 0.439453125, - "learning_rate": 3.080418372560768e-05, - "loss": 1.971, + "epoch": 0.5859012651465482, + "grad_norm": 0.64453125, + "learning_rate": 8.748014034535983e-05, + "loss": 2.3165, "step": 1508 }, { - "epoch": 0.48326661329063253, - "grad_norm": 1.4765625, - "learning_rate": 3.077698402325857e-05, - "loss": 2.0729, + "epoch": 0.5862897938369637, + "grad_norm": 0.65234375, + "learning_rate": 8.734551374388436e-05, + "loss": 2.3892, "step": 1509 }, { - "epoch": 0.4835868694955965, - "grad_norm": 0.609375, - "learning_rate": 3.0749777094918924e-05, - "loss": 2.444, + "epoch": 0.5866783225273792, + "grad_norm": 0.67578125, + "learning_rate": 8.72109104470386e-05, + "loss": 2.3335, "step": 1510 }, { - "epoch": 0.48390712570056044, - "grad_norm": 0.44140625, - "learning_rate": 3.072256297461983e-05, - "loss": 2.6036, + "epoch": 0.5870668512177946, + "grad_norm": 0.72265625, + "learning_rate": 8.707633070270932e-05, + "loss": 2.4179, "step": 1511 }, { - "epoch": 0.48422738190552445, - "grad_norm": 0.4921875, - "learning_rate": 3.069534169640136e-05, - "loss": 2.5759, + "epoch": 0.5874553799082101, + "grad_norm": 0.6796875, + "learning_rate": 8.694177475873997e-05, + "loss": 2.3839, "step": 1512 }, { - "epoch": 0.4845476381104884, - "grad_norm": 0.484375, - "learning_rate": 3.066811329431254e-05, - "loss": 2.5265, + "epoch": 0.5878439085986256, + "grad_norm": 0.69140625, + "learning_rate": 8.680724286293008e-05, + "loss": 2.3786, "step": 1513 }, { - "epoch": 0.48486789431545235, - "grad_norm": 0.6875, - "learning_rate": 3.0640877802411316e-05, - "loss": 2.5761, + "epoch": 0.5882324372890411, + "grad_norm": 0.6796875, + "learning_rate": 8.667273526303497e-05, + "loss": 2.3887, "step": 1514 }, { - "epoch": 0.4851881505204163, - "grad_norm": 0.478515625, - "learning_rate": 3.0613635254764495e-05, - "loss": 1.9635, + "epoch": 0.5886209659794566, + "grad_norm": 0.71484375, + "learning_rate": 8.653825220676521e-05, + "loss": 2.461, "step": 1515 }, { - "epoch": 0.4855084067253803, - "grad_norm": 0.435546875, - "learning_rate": 3.058638568544772e-05, - "loss": 1.9067, + "epoch": 0.5890094946698721, + "grad_norm": 0.69140625, + "learning_rate": 8.640379394178617e-05, + "loss": 2.3804, "step": 1516 }, { - "epoch": 0.48582866293034427, - "grad_norm": 0.59375, - "learning_rate": 3.055912912854538e-05, - "loss": 2.7006, + "epoch": 0.5893980233602875, + "grad_norm": 0.66015625, + "learning_rate": 8.62693607157175e-05, + "loss": 2.3552, "step": 1517 }, { - "epoch": 0.4861489191353082, - "grad_norm": 1.8671875, - "learning_rate": 3.0531865618150654e-05, - "loss": 2.7706, + "epoch": 0.589786552050703, + "grad_norm": 0.6875, + "learning_rate": 8.613495277613281e-05, + "loss": 2.3252, "step": 1518 }, { - "epoch": 0.48646917534027223, - "grad_norm": 2.703125, - "learning_rate": 3.050459518836539e-05, - "loss": 2.2856, + "epoch": 0.5901750807411185, + "grad_norm": 0.6484375, + "learning_rate": 8.600057037055914e-05, + "loss": 2.3619, "step": 1519 }, { - "epoch": 0.4867894315452362, - "grad_norm": 0.498046875, - "learning_rate": 3.0477317873300103e-05, - "loss": 2.2043, + "epoch": 0.590563609431534, + "grad_norm": 0.6953125, + "learning_rate": 8.586621374647646e-05, + "loss": 2.3739, "step": 1520 }, { - "epoch": 0.48710968775020014, - "grad_norm": 1.3828125, - "learning_rate": 3.045003370707391e-05, - "loss": 1.7819, + "epoch": 0.5909521381219495, + "grad_norm": 0.80078125, + "learning_rate": 8.573188315131728e-05, + "loss": 2.4229, "step": 1521 }, { - "epoch": 0.48742994395516415, - "grad_norm": 0.57421875, - "learning_rate": 3.0422742723814503e-05, - "loss": 2.2534, + "epoch": 0.591340666812365, + "grad_norm": 0.6484375, + "learning_rate": 8.559757883246617e-05, + "loss": 2.3038, "step": 1522 }, { - "epoch": 0.4877502001601281, - "grad_norm": 0.37109375, - "learning_rate": 3.0395444957658097e-05, - "loss": 1.9324, + "epoch": 0.5917291955027804, + "grad_norm": 0.68359375, + "learning_rate": 8.546330103725937e-05, + "loss": 2.3077, "step": 1523 }, { - "epoch": 0.48807045636509205, - "grad_norm": 0.498046875, - "learning_rate": 3.0368140442749398e-05, - "loss": 2.167, + "epoch": 0.5921177241931959, + "grad_norm": 0.64453125, + "learning_rate": 8.53290500129842e-05, + "loss": 2.3405, "step": 1524 }, { - "epoch": 0.48839071257005606, - "grad_norm": 0.546875, - "learning_rate": 3.034082921324155e-05, - "loss": 2.6054, + "epoch": 0.5925062528836114, + "grad_norm": 0.68359375, + "learning_rate": 8.51948260068787e-05, + "loss": 2.3246, "step": 1525 }, { - "epoch": 0.48871096877502, - "grad_norm": 0.546875, - "learning_rate": 3.03135113032961e-05, - "loss": 2.4174, + "epoch": 0.5928947815740269, + "grad_norm": 0.6796875, + "learning_rate": 8.506062926613112e-05, + "loss": 2.4295, "step": 1526 }, { - "epoch": 0.48903122497998397, - "grad_norm": 0.62890625, - "learning_rate": 3.0286186747082934e-05, - "loss": 2.3323, + "epoch": 0.5932833102644424, + "grad_norm": 0.63671875, + "learning_rate": 8.492646003787958e-05, + "loss": 2.3574, "step": 1527 }, { - "epoch": 0.489351481184948, - "grad_norm": 1.953125, - "learning_rate": 3.0258855578780265e-05, - "loss": 2.1546, + "epoch": 0.5936718389548579, + "grad_norm": 0.69140625, + "learning_rate": 8.479231856921148e-05, + "loss": 2.4245, "step": 1528 }, { - "epoch": 0.48967173738991193, - "grad_norm": 0.55078125, - "learning_rate": 3.023151783257459e-05, - "loss": 2.4691, + "epoch": 0.5940603676452733, + "grad_norm": 0.671875, + "learning_rate": 8.465820510716311e-05, + "loss": 2.3591, "step": 1529 }, { - "epoch": 0.4899919935948759, - "grad_norm": 0.396484375, - "learning_rate": 3.0204173542660613e-05, - "loss": 2.0822, + "epoch": 0.5944488963356888, + "grad_norm": 0.7109375, + "learning_rate": 8.452411989871916e-05, + "loss": 2.4629, "step": 1530 }, { - "epoch": 0.4903122497998399, - "grad_norm": 0.89453125, - "learning_rate": 3.0176822743241222e-05, - "loss": 2.6899, + "epoch": 0.5948374250261043, + "grad_norm": 0.66015625, + "learning_rate": 8.439006319081228e-05, + "loss": 2.3193, "step": 1531 }, { - "epoch": 0.49063250600480385, - "grad_norm": 0.51171875, - "learning_rate": 3.014946546852746e-05, - "loss": 2.5449, + "epoch": 0.5952259537165198, + "grad_norm": 0.69921875, + "learning_rate": 8.425603523032269e-05, + "loss": 2.3394, "step": 1532 }, { - "epoch": 0.4909527622097678, - "grad_norm": 0.71484375, - "learning_rate": 3.0122101752738468e-05, - "loss": 2.0894, + "epoch": 0.5956144824069353, + "grad_norm": 0.65234375, + "learning_rate": 8.412203626407766e-05, + "loss": 2.322, "step": 1533 }, { - "epoch": 0.4912730184147318, - "grad_norm": 0.46484375, - "learning_rate": 3.0094731630101436e-05, - "loss": 2.2903, + "epoch": 0.5960030110973508, + "grad_norm": 0.6796875, + "learning_rate": 8.398806653885098e-05, + "loss": 2.3262, "step": 1534 }, { - "epoch": 0.49159327461969576, - "grad_norm": 0.427734375, - "learning_rate": 3.0067355134851565e-05, - "loss": 2.2904, + "epoch": 0.5963915397877662, + "grad_norm": 0.69921875, + "learning_rate": 8.385412630136267e-05, + "loss": 2.3941, "step": 1535 }, { - "epoch": 0.4919135308246597, - "grad_norm": 0.365234375, - "learning_rate": 3.0039972301232045e-05, - "loss": 2.2971, + "epoch": 0.5967800684781817, + "grad_norm": 0.73046875, + "learning_rate": 8.372021579827844e-05, + "loss": 2.3708, "step": 1536 }, { - "epoch": 0.4922337870296237, - "grad_norm": 0.46875, - "learning_rate": 3.001258316349398e-05, - "loss": 1.7656, + "epoch": 0.5971685971685972, + "grad_norm": 0.7421875, + "learning_rate": 8.358633527620923e-05, + "loss": 2.3973, "step": 1537 }, { - "epoch": 0.4925540432345877, - "grad_norm": 0.48828125, - "learning_rate": 2.998518775589635e-05, - "loss": 2.0205, + "epoch": 0.5975571258590127, + "grad_norm": 15.1875, + "learning_rate": 8.345248498171073e-05, + "loss": 2.3555, "step": 1538 }, { - "epoch": 0.49287429943955163, - "grad_norm": 0.59765625, - "learning_rate": 2.9957786112706006e-05, - "loss": 2.3737, + "epoch": 0.5979456545494282, + "grad_norm": 0.765625, + "learning_rate": 8.331866516128302e-05, + "loss": 2.3586, "step": 1539 }, { - "epoch": 0.4931945556445156, - "grad_norm": 0.40625, - "learning_rate": 2.9930378268197577e-05, - "loss": 2.1592, + "epoch": 0.5983341832398437, + "grad_norm": 0.74609375, + "learning_rate": 8.318487606137001e-05, + "loss": 2.3895, "step": 1540 }, { - "epoch": 0.4935148118494796, - "grad_norm": 0.5390625, - "learning_rate": 2.990296425665345e-05, - "loss": 2.3567, + "epoch": 0.598722711930259, + "grad_norm": 0.70703125, + "learning_rate": 8.305111792835911e-05, + "loss": 2.341, "step": 1541 }, { - "epoch": 0.49383506805444355, - "grad_norm": 0.53515625, - "learning_rate": 2.987554411236373e-05, - "loss": 2.3767, + "epoch": 0.5991112406206746, + "grad_norm": 0.63671875, + "learning_rate": 8.291739100858058e-05, + "loss": 2.3371, "step": 1542 }, { - "epoch": 0.4941553242594075, - "grad_norm": 1.625, - "learning_rate": 2.9848117869626192e-05, - "loss": 2.5746, + "epoch": 0.5994997693110901, + "grad_norm": 0.65234375, + "learning_rate": 8.278369554830729e-05, + "loss": 2.3742, "step": 1543 }, { - "epoch": 0.4944755804643715, - "grad_norm": 0.349609375, - "learning_rate": 2.9820685562746254e-05, - "loss": 1.4638, + "epoch": 0.5998882980015056, + "grad_norm": 0.6640625, + "learning_rate": 8.265003179375418e-05, + "loss": 2.394, "step": 1544 }, { - "epoch": 0.49479583666933546, - "grad_norm": 0.97265625, - "learning_rate": 2.9793247226036885e-05, - "loss": 1.9633, + "epoch": 0.6002768266919211, + "grad_norm": 0.70703125, + "learning_rate": 8.251639999107776e-05, + "loss": 2.3719, "step": 1545 }, { - "epoch": 0.4951160928742994, - "grad_norm": 0.4765625, - "learning_rate": 2.9765802893818627e-05, - "loss": 1.7408, + "epoch": 0.6006653553823366, + "grad_norm": 0.71875, + "learning_rate": 8.238280038637566e-05, + "loss": 2.3728, "step": 1546 }, { - "epoch": 0.4954363490792634, - "grad_norm": 0.474609375, - "learning_rate": 2.9738352600419506e-05, - "loss": 2.0108, + "epoch": 0.601053884072752, + "grad_norm": 0.68359375, + "learning_rate": 8.224923322568637e-05, + "loss": 2.3766, "step": 1547 }, { - "epoch": 0.4957566052842274, - "grad_norm": 0.388671875, - "learning_rate": 2.9710896380175024e-05, - "loss": 2.3077, + "epoch": 0.6014424127631675, + "grad_norm": 0.66015625, + "learning_rate": 8.211569875498837e-05, + "loss": 2.3564, "step": 1548 }, { - "epoch": 0.49607686148919133, - "grad_norm": 0.86328125, - "learning_rate": 2.9683434267428055e-05, - "loss": 2.0489, + "epoch": 0.601830941453583, + "grad_norm": 0.6640625, + "learning_rate": 8.198219722020019e-05, + "loss": 2.3739, "step": 1549 }, { - "epoch": 0.49639711769415534, - "grad_norm": 1.3125, - "learning_rate": 2.9655966296528893e-05, - "loss": 2.8007, + "epoch": 0.6022194701439985, + "grad_norm": 0.71484375, + "learning_rate": 8.184872886717956e-05, + "loss": 2.4159, "step": 1550 }, { - "epoch": 0.4967173738991193, - "grad_norm": 0.412109375, - "learning_rate": 2.962849250183513e-05, - "loss": 2.0005, + "epoch": 0.602607998834414, + "grad_norm": 0.734375, + "learning_rate": 8.17152939417232e-05, + "loss": 2.3322, "step": 1551 }, { - "epoch": 0.49703763010408325, - "grad_norm": 0.6953125, - "learning_rate": 2.9601012917711646e-05, - "loss": 1.5309, + "epoch": 0.6029965275248295, + "grad_norm": 0.69921875, + "learning_rate": 8.158189268956619e-05, + "loss": 2.3703, "step": 1552 }, { - "epoch": 0.49735788630904726, - "grad_norm": 0.46484375, - "learning_rate": 2.957352757853056e-05, - "loss": 2.4253, + "epoch": 0.6033850562152449, + "grad_norm": 0.65625, + "learning_rate": 8.144852535638162e-05, + "loss": 2.3311, "step": 1553 }, { - "epoch": 0.4976781425140112, - "grad_norm": 0.416015625, - "learning_rate": 2.9546036518671206e-05, - "loss": 1.8466, + "epoch": 0.6037735849056604, + "grad_norm": 0.6875, + "learning_rate": 8.131519218778008e-05, + "loss": 2.3832, "step": 1554 }, { - "epoch": 0.49799839871897517, - "grad_norm": 0.609375, - "learning_rate": 2.9518539772520064e-05, - "loss": 2.2555, + "epoch": 0.6041621135960759, + "grad_norm": 0.640625, + "learning_rate": 8.118189342930936e-05, + "loss": 2.3851, "step": 1555 }, { - "epoch": 0.4983186549239392, - "grad_norm": 0.419921875, - "learning_rate": 2.9491037374470715e-05, - "loss": 2.0387, + "epoch": 0.6045506422864914, + "grad_norm": 0.703125, + "learning_rate": 8.104862932645374e-05, + "loss": 2.3566, "step": 1556 }, { - "epoch": 0.49863891112890313, - "grad_norm": 0.7578125, - "learning_rate": 2.9463529358923824e-05, - "loss": 2.4217, + "epoch": 0.6049391709769069, + "grad_norm": 0.671875, + "learning_rate": 8.091540012463375e-05, + "loss": 2.3132, "step": 1557 }, { - "epoch": 0.4989591673338671, - "grad_norm": 0.640625, - "learning_rate": 2.9436015760287082e-05, - "loss": 2.1835, + "epoch": 0.6053276996673224, + "grad_norm": 0.66015625, + "learning_rate": 8.078220606920564e-05, + "loss": 2.3965, "step": 1558 }, { - "epoch": 0.4992794235388311, - "grad_norm": 0.51953125, - "learning_rate": 2.9408496612975156e-05, - "loss": 2.1866, + "epoch": 0.6057162283577378, + "grad_norm": 0.72265625, + "learning_rate": 8.064904740546092e-05, + "loss": 2.4171, "step": 1559 }, { - "epoch": 0.49959967974379504, - "grad_norm": 0.4296875, - "learning_rate": 2.9380971951409652e-05, - "loss": 2.0304, + "epoch": 0.6061047570481533, + "grad_norm": 0.6875, + "learning_rate": 8.051592437862597e-05, + "loss": 2.3941, "step": 1560 }, { - "epoch": 0.499919935948759, - "grad_norm": 0.640625, - "learning_rate": 2.9353441810019093e-05, - "loss": 1.4819, + "epoch": 0.6064932857385688, + "grad_norm": 0.69921875, + "learning_rate": 8.038283723386144e-05, + "loss": 2.4175, "step": 1561 }, { - "epoch": 0.500240192153723, - "grad_norm": 0.65234375, - "learning_rate": 2.9325906223238836e-05, - "loss": 1.8902, + "epoch": 0.6068818144289843, + "grad_norm": 0.6640625, + "learning_rate": 8.024978621626199e-05, + "loss": 2.3083, "step": 1562 }, { - "epoch": 0.500560448358687, - "grad_norm": 0.419921875, - "learning_rate": 2.929836522551104e-05, - "loss": 2.3322, + "epoch": 0.6072703431193998, + "grad_norm": 0.70703125, + "learning_rate": 8.011677157085571e-05, + "loss": 2.3282, "step": 1563 }, { - "epoch": 0.500880704563651, - "grad_norm": 0.470703125, - "learning_rate": 2.927081885128467e-05, - "loss": 1.9023, + "epoch": 0.6076588718098153, + "grad_norm": 0.68359375, + "learning_rate": 7.998379354260375e-05, + "loss": 2.3535, "step": 1564 }, { - "epoch": 0.5012009607686149, - "grad_norm": 0.4296875, - "learning_rate": 2.9243267135015383e-05, - "loss": 2.1684, + "epoch": 0.6080474005002307, + "grad_norm": 0.71875, + "learning_rate": 7.985085237639973e-05, + "loss": 2.424, "step": 1565 }, { - "epoch": 0.5015212169735789, - "grad_norm": 0.54296875, - "learning_rate": 2.921571011116552e-05, - "loss": 2.0673, + "epoch": 0.6084359291906462, + "grad_norm": 0.734375, + "learning_rate": 7.971794831706947e-05, + "loss": 2.2892, "step": 1566 }, { - "epoch": 0.5018414731785429, - "grad_norm": 0.50390625, - "learning_rate": 2.9188147814204086e-05, - "loss": 2.5705, + "epoch": 0.6088244578810617, + "grad_norm": 0.7578125, + "learning_rate": 7.958508160937044e-05, + "loss": 2.315, "step": 1567 }, { - "epoch": 0.5021617293835068, - "grad_norm": 0.515625, - "learning_rate": 2.916058027860665e-05, - "loss": 2.6609, + "epoch": 0.6092129865714772, + "grad_norm": 0.7265625, + "learning_rate": 7.94522524979913e-05, + "loss": 2.3734, "step": 1568 }, { - "epoch": 0.5024819855884708, - "grad_norm": 0.484375, - "learning_rate": 2.913300753885536e-05, - "loss": 2.2706, + "epoch": 0.6096015152618927, + "grad_norm": 0.6640625, + "learning_rate": 7.931946122755149e-05, + "loss": 2.333, "step": 1569 }, { - "epoch": 0.5028022417934348, - "grad_norm": 0.50390625, - "learning_rate": 2.910542962943885e-05, - "loss": 2.5645, + "epoch": 0.6099900439523082, + "grad_norm": 0.6640625, + "learning_rate": 7.918670804260074e-05, + "loss": 2.3811, "step": 1570 }, { - "epoch": 0.5031224979983987, - "grad_norm": 0.578125, - "learning_rate": 2.907784658485225e-05, - "loss": 2.7824, + "epoch": 0.6103785726427235, + "grad_norm": 0.66796875, + "learning_rate": 7.905399318761869e-05, + "loss": 2.362, "step": 1571 }, { - "epoch": 0.5034427542033627, - "grad_norm": 0.58203125, - "learning_rate": 2.905025843959708e-05, - "loss": 1.5037, + "epoch": 0.610767101333139, + "grad_norm": 0.66015625, + "learning_rate": 7.89213169070143e-05, + "loss": 2.3697, "step": 1572 }, { - "epoch": 0.5037630104083266, - "grad_norm": 1.1015625, - "learning_rate": 2.902266522818125e-05, - "loss": 2.2413, + "epoch": 0.6111556300235546, + "grad_norm": 0.65234375, + "learning_rate": 7.878867944512561e-05, + "loss": 2.3693, "step": 1573 }, { - "epoch": 0.5040832666132906, - "grad_norm": 0.439453125, - "learning_rate": 2.899506698511903e-05, - "loss": 2.5246, + "epoch": 0.61154415871397, + "grad_norm": 0.671875, + "learning_rate": 7.865608104621907e-05, + "loss": 2.3939, "step": 1574 }, { - "epoch": 0.5044035228182546, - "grad_norm": 0.4375, - "learning_rate": 2.8967463744930958e-05, - "loss": 1.7726, + "epoch": 0.6119326874043856, + "grad_norm": 0.65234375, + "learning_rate": 7.852352195448923e-05, + "loss": 2.4086, "step": 1575 }, { - "epoch": 0.5047237790232185, - "grad_norm": 0.51953125, - "learning_rate": 2.8939855542143828e-05, - "loss": 1.8297, + "epoch": 0.612321216094801, + "grad_norm": 0.69921875, + "learning_rate": 7.839100241405828e-05, + "loss": 2.3811, "step": 1576 }, { - "epoch": 0.5050440352281825, - "grad_norm": 0.4453125, - "learning_rate": 2.8912242411290646e-05, - "loss": 1.8675, + "epoch": 0.6127097447852164, + "grad_norm": 0.6328125, + "learning_rate": 7.825852266897553e-05, + "loss": 2.2625, "step": 1577 }, { - "epoch": 0.5053642914331465, - "grad_norm": 1.0546875, - "learning_rate": 2.8884624386910586e-05, - "loss": 2.3941, + "epoch": 0.613098273475632, + "grad_norm": 0.66796875, + "learning_rate": 7.8126082963217e-05, + "loss": 2.4848, "step": 1578 }, { - "epoch": 0.5056845476381104, - "grad_norm": 0.51171875, - "learning_rate": 2.8857001503548925e-05, - "loss": 2.5356, + "epoch": 0.6134868021660475, + "grad_norm": 0.72265625, + "learning_rate": 7.799368354068502e-05, + "loss": 2.4191, "step": 1579 }, { - "epoch": 0.5060048038430744, - "grad_norm": 0.6640625, - "learning_rate": 2.882937379575704e-05, - "loss": 2.5708, + "epoch": 0.613875330856463, + "grad_norm": 0.6484375, + "learning_rate": 7.786132464520768e-05, + "loss": 2.3593, "step": 1580 }, { - "epoch": 0.5063250600480385, - "grad_norm": 0.50390625, - "learning_rate": 2.8801741298092332e-05, - "loss": 2.1658, + "epoch": 0.6142638595468785, + "grad_norm": 0.67578125, + "learning_rate": 7.772900652053848e-05, + "loss": 2.3714, "step": 1581 }, { - "epoch": 0.5066453162530024, - "grad_norm": 0.50390625, - "learning_rate": 2.8774104045118183e-05, - "loss": 1.7119, + "epoch": 0.6146523882372938, + "grad_norm": 0.6875, + "learning_rate": 7.759672941035579e-05, + "loss": 2.3637, "step": 1582 }, { - "epoch": 0.5069655724579664, - "grad_norm": 0.71875, - "learning_rate": 2.8746462071403952e-05, - "loss": 1.936, + "epoch": 0.6150409169277093, + "grad_norm": 0.6328125, + "learning_rate": 7.746449355826246e-05, + "loss": 2.2919, "step": 1583 }, { - "epoch": 0.5072858286629304, - "grad_norm": 0.462890625, - "learning_rate": 2.8718815411524863e-05, - "loss": 2.1219, + "epoch": 0.6154294456181248, + "grad_norm": 0.6796875, + "learning_rate": 7.733229920778542e-05, + "loss": 2.3922, "step": 1584 }, { - "epoch": 0.5076060848678943, - "grad_norm": 0.52734375, - "learning_rate": 2.8691164100062034e-05, - "loss": 2.5705, + "epoch": 0.6158179743085404, + "grad_norm": 0.64453125, + "learning_rate": 7.720014660237512e-05, + "loss": 2.4041, "step": 1585 }, { - "epoch": 0.5079263410728583, - "grad_norm": 0.48046875, - "learning_rate": 2.8663508171602386e-05, - "loss": 2.2458, + "epoch": 0.6162065029989559, + "grad_norm": 0.65625, + "learning_rate": 7.706803598540506e-05, + "loss": 2.3844, "step": 1586 }, { - "epoch": 0.5082465972778223, - "grad_norm": 1.390625, - "learning_rate": 2.8635847660738635e-05, - "loss": 1.8396, + "epoch": 0.6165950316893714, + "grad_norm": 0.671875, + "learning_rate": 7.693596760017156e-05, + "loss": 2.4035, "step": 1587 }, { - "epoch": 0.5085668534827862, - "grad_norm": 0.515625, - "learning_rate": 2.8608182602069195e-05, - "loss": 2.5934, + "epoch": 0.6169835603797867, + "grad_norm": 0.67578125, + "learning_rate": 7.680394168989306e-05, + "loss": 2.3288, "step": 1588 }, { - "epoch": 0.5088871096877502, - "grad_norm": 0.625, - "learning_rate": 2.8580513030198186e-05, - "loss": 2.4622, + "epoch": 0.6173720890702022, + "grad_norm": 0.6796875, + "learning_rate": 7.66719584977098e-05, + "loss": 2.3953, "step": 1589 }, { - "epoch": 0.5092073658927142, - "grad_norm": 0.490234375, - "learning_rate": 2.8552838979735393e-05, - "loss": 2.2395, + "epoch": 0.6177606177606177, + "grad_norm": 0.6640625, + "learning_rate": 7.654001826668342e-05, + "loss": 2.3341, "step": 1590 }, { - "epoch": 0.5095276220976781, - "grad_norm": 0.51953125, - "learning_rate": 2.8525160485296167e-05, - "loss": 2.172, + "epoch": 0.6181491464510332, + "grad_norm": 0.66796875, + "learning_rate": 7.640812123979635e-05, + "loss": 2.3844, "step": 1591 }, { - "epoch": 0.5098478783026421, - "grad_norm": 0.5078125, - "learning_rate": 2.849747758150143e-05, - "loss": 2.4499, + "epoch": 0.6185376751414488, + "grad_norm": 0.6328125, + "learning_rate": 7.627626765995144e-05, + "loss": 2.3106, "step": 1592 }, { - "epoch": 0.5101681345076061, - "grad_norm": 0.5390625, - "learning_rate": 2.846979030297764e-05, - "loss": 2.2869, + "epoch": 0.6189262038318643, + "grad_norm": 0.76171875, + "learning_rate": 7.61444577699716e-05, + "loss": 2.3895, "step": 1593 }, { - "epoch": 0.51048839071257, - "grad_norm": 0.37109375, - "learning_rate": 2.8442098684356707e-05, - "loss": 1.8278, + "epoch": 0.6193147325222796, + "grad_norm": 0.6796875, + "learning_rate": 7.601269181259926e-05, + "loss": 2.3381, "step": 1594 }, { - "epoch": 0.510808646917534, - "grad_norm": 0.458984375, - "learning_rate": 2.841440276027596e-05, - "loss": 2.2451, + "epoch": 0.6197032612126951, + "grad_norm": 0.63671875, + "learning_rate": 7.588097003049596e-05, + "loss": 2.3945, "step": 1595 }, { - "epoch": 0.511128903122498, - "grad_norm": 0.859375, - "learning_rate": 2.8386702565378137e-05, - "loss": 2.3676, + "epoch": 0.6200917899031106, + "grad_norm": 0.640625, + "learning_rate": 7.574929266624185e-05, + "loss": 2.3523, "step": 1596 }, { - "epoch": 0.5114491593274619, - "grad_norm": 0.921875, - "learning_rate": 2.8358998134311316e-05, - "loss": 2.4343, + "epoch": 0.6204803185935261, + "grad_norm": 0.66015625, + "learning_rate": 7.561765996233527e-05, + "loss": 2.3788, "step": 1597 }, { - "epoch": 0.5117694155324259, - "grad_norm": 0.62109375, - "learning_rate": 2.8331289501728854e-05, - "loss": 2.516, + "epoch": 0.6208688472839416, + "grad_norm": 0.6875, + "learning_rate": 7.548607216119236e-05, + "loss": 2.426, "step": 1598 }, { - "epoch": 0.51208967173739, - "grad_norm": 0.5234375, - "learning_rate": 2.830357670228937e-05, - "loss": 2.5648, + "epoch": 0.6212573759743572, + "grad_norm": 0.625, + "learning_rate": 7.535452950514655e-05, + "loss": 2.2676, "step": 1599 }, { - "epoch": 0.5124099279423538, - "grad_norm": 0.5234375, - "learning_rate": 2.8275859770656714e-05, - "loss": 2.4266, + "epoch": 0.6216459046647725, + "grad_norm": 0.75390625, + "learning_rate": 7.522303223644814e-05, + "loss": 2.3779, "step": 1600 }, { - "epoch": 0.5127301841473179, - "grad_norm": 0.43359375, - "learning_rate": 2.8248138741499885e-05, - "loss": 2.4382, + "epoch": 0.622034433355188, + "grad_norm": 0.62890625, + "learning_rate": 7.509158059726375e-05, + "loss": 2.299, "step": 1601 }, { - "epoch": 0.5130504403522819, - "grad_norm": 0.578125, - "learning_rate": 2.822041364949301e-05, - "loss": 2.4932, + "epoch": 0.6224229620456035, + "grad_norm": 0.68359375, + "learning_rate": 7.496017482967613e-05, + "loss": 2.3966, "step": 1602 }, { - "epoch": 0.5133706965572458, - "grad_norm": 0.51171875, - "learning_rate": 2.8192684529315305e-05, - "loss": 2.121, + "epoch": 0.622811490736019, + "grad_norm": 0.6796875, + "learning_rate": 7.482881517568343e-05, + "loss": 2.3278, "step": 1603 }, { - "epoch": 0.5136909527622098, - "grad_norm": 0.5859375, - "learning_rate": 2.816495141565102e-05, - "loss": 2.0563, + "epoch": 0.6232000194264345, + "grad_norm": 0.703125, + "learning_rate": 7.469750187719895e-05, + "loss": 2.3683, "step": 1604 }, { - "epoch": 0.5140112089671738, - "grad_norm": 0.396484375, - "learning_rate": 2.8137214343189395e-05, - "loss": 2.4438, + "epoch": 0.62358854811685, + "grad_norm": 0.66796875, + "learning_rate": 7.456623517605051e-05, + "loss": 2.3648, "step": 1605 }, { - "epoch": 0.5143314651721377, - "grad_norm": 0.369140625, - "learning_rate": 2.8109473346624627e-05, - "loss": 2.3331, + "epoch": 0.6239770768072654, + "grad_norm": 0.671875, + "learning_rate": 7.443501531398027e-05, + "loss": 2.3241, "step": 1606 }, { - "epoch": 0.5146517213771017, - "grad_norm": 0.43359375, - "learning_rate": 2.808172846065583e-05, - "loss": 2.2871, + "epoch": 0.6243656054976809, + "grad_norm": 0.640625, + "learning_rate": 7.430384253264402e-05, + "loss": 2.3535, "step": 1607 }, { - "epoch": 0.5149719775820657, - "grad_norm": 0.53515625, - "learning_rate": 2.8053979719986972e-05, - "loss": 2.4588, + "epoch": 0.6247541341880964, + "grad_norm": 0.6796875, + "learning_rate": 7.41727170736109e-05, + "loss": 2.3948, "step": 1608 }, { - "epoch": 0.5152922337870296, - "grad_norm": 0.5703125, - "learning_rate": 2.802622715932684e-05, - "loss": 2.5462, + "epoch": 0.625142662878512, + "grad_norm": 0.62890625, + "learning_rate": 7.404163917836284e-05, + "loss": 2.3515, "step": 1609 }, { - "epoch": 0.5156124899919936, - "grad_norm": 0.4765625, - "learning_rate": 2.799847081338901e-05, - "loss": 1.7209, + "epoch": 0.6255311915689274, + "grad_norm": 0.67578125, + "learning_rate": 7.391060908829426e-05, + "loss": 2.3847, "step": 1610 }, { - "epoch": 0.5159327461969576, - "grad_norm": 1.4609375, - "learning_rate": 2.7970710716891783e-05, - "loss": 2.7945, + "epoch": 0.625919720259343, + "grad_norm": 0.67578125, + "learning_rate": 7.377962704471147e-05, + "loss": 2.4048, "step": 1611 }, { - "epoch": 0.5162530024019215, - "grad_norm": 0.5390625, - "learning_rate": 2.7942946904558155e-05, - "loss": 2.2602, + "epoch": 0.6263082489497583, + "grad_norm": 0.66796875, + "learning_rate": 7.364869328883232e-05, + "loss": 2.4245, "step": 1612 }, { - "epoch": 0.5165732586068855, - "grad_norm": 0.466796875, - "learning_rate": 2.791517941111577e-05, - "loss": 2.0592, + "epoch": 0.6266967776401738, + "grad_norm": 0.765625, + "learning_rate": 7.351780806178577e-05, + "loss": 2.3432, "step": 1613 }, { - "epoch": 0.5168935148118495, - "grad_norm": 0.5234375, - "learning_rate": 2.7887408271296883e-05, - "loss": 2.1679, + "epoch": 0.6270853063305893, + "grad_norm": 0.63671875, + "learning_rate": 7.338697160461135e-05, + "loss": 2.425, "step": 1614 }, { - "epoch": 0.5172137710168134, - "grad_norm": 0.52734375, - "learning_rate": 2.7859633519838296e-05, - "loss": 2.5147, + "epoch": 0.6274738350210048, + "grad_norm": 0.65234375, + "learning_rate": 7.325618415825878e-05, + "loss": 2.2868, "step": 1615 }, { - "epoch": 0.5175340272217774, - "grad_norm": 0.62109375, - "learning_rate": 2.7831855191481336e-05, - "loss": 2.471, + "epoch": 0.6278623637114203, + "grad_norm": 0.74609375, + "learning_rate": 7.31254459635876e-05, + "loss": 2.3369, "step": 1616 }, { - "epoch": 0.5178542834267414, - "grad_norm": 0.69921875, - "learning_rate": 2.7804073320971812e-05, - "loss": 2.2137, + "epoch": 0.6282508924018358, + "grad_norm": 0.6640625, + "learning_rate": 7.299475726136655e-05, + "loss": 2.3966, "step": 1617 }, { - "epoch": 0.5181745396317053, - "grad_norm": 1.640625, - "learning_rate": 2.777628794305995e-05, - "loss": 2.0848, + "epoch": 0.6286394210922512, + "grad_norm": 0.6328125, + "learning_rate": 7.286411829227326e-05, + "loss": 2.3717, "step": 1618 }, { - "epoch": 0.5184947958366694, - "grad_norm": 0.53515625, - "learning_rate": 2.774849909250037e-05, - "loss": 2.0605, + "epoch": 0.6290279497826667, + "grad_norm": 0.640625, + "learning_rate": 7.27335292968938e-05, + "loss": 2.3432, "step": 1619 }, { - "epoch": 0.5188150520416333, - "grad_norm": 0.75, - "learning_rate": 2.7720706804052026e-05, - "loss": 2.6878, + "epoch": 0.6294164784730822, + "grad_norm": 0.66015625, + "learning_rate": 7.260299051572216e-05, + "loss": 2.3927, "step": 1620 }, { - "epoch": 0.5191353082465973, - "grad_norm": 0.490234375, - "learning_rate": 2.7692911112478193e-05, - "loss": 2.4631, + "epoch": 0.6298050071634977, + "grad_norm": 0.71875, + "learning_rate": 7.24725021891599e-05, + "loss": 2.3983, "step": 1621 }, { - "epoch": 0.5194555644515613, - "grad_norm": 0.62109375, - "learning_rate": 2.7665112052546387e-05, - "loss": 2.5574, + "epoch": 0.6301935358539132, + "grad_norm": 0.6484375, + "learning_rate": 7.234206455751562e-05, + "loss": 2.2967, "step": 1622 }, { - "epoch": 0.5197758206565252, - "grad_norm": 0.421875, - "learning_rate": 2.763730965902834e-05, - "loss": 2.0565, + "epoch": 0.6305820645443287, + "grad_norm": 0.69140625, + "learning_rate": 7.221167786100458e-05, + "loss": 2.4191, "step": 1623 }, { - "epoch": 0.5200960768614892, - "grad_norm": 0.462890625, - "learning_rate": 2.7609503966699945e-05, - "loss": 1.9217, + "epoch": 0.6309705932347441, + "grad_norm": 0.67578125, + "learning_rate": 7.208134233974825e-05, + "loss": 2.3658, "step": 1624 }, { - "epoch": 0.5204163330664532, - "grad_norm": 0.46875, - "learning_rate": 2.7581695010341246e-05, - "loss": 2.049, + "epoch": 0.6313591219251596, + "grad_norm": 0.62890625, + "learning_rate": 7.195105823377384e-05, + "loss": 2.3469, "step": 1625 }, { - "epoch": 0.5207365892714171, - "grad_norm": 0.439453125, - "learning_rate": 2.7553882824736348e-05, - "loss": 2.1618, + "epoch": 0.6317476506155751, + "grad_norm": 0.65234375, + "learning_rate": 7.182082578301388e-05, + "loss": 2.3892, "step": 1626 }, { - "epoch": 0.5210568454763811, - "grad_norm": 0.494140625, - "learning_rate": 2.7526067444673402e-05, - "loss": 1.9671, + "epoch": 0.6321361793059906, + "grad_norm": 0.64453125, + "learning_rate": 7.169064522730573e-05, + "loss": 2.3862, "step": 1627 }, { - "epoch": 0.5213771016813451, - "grad_norm": 0.60546875, - "learning_rate": 2.749824890494455e-05, - "loss": 2.1159, + "epoch": 0.6325247079964061, + "grad_norm": 0.6171875, + "learning_rate": 7.156051680639126e-05, + "loss": 2.3799, "step": 1628 }, { - "epoch": 0.521697357886309, - "grad_norm": 0.423828125, - "learning_rate": 2.74704272403459e-05, - "loss": 1.5528, + "epoch": 0.6329132366868216, + "grad_norm": 0.6796875, + "learning_rate": 7.143044075991626e-05, + "loss": 2.3772, "step": 1629 }, { - "epoch": 0.522017614091273, - "grad_norm": 0.98828125, - "learning_rate": 2.744260248567745e-05, - "loss": 2.162, + "epoch": 0.633301765377237, + "grad_norm": 0.65625, + "learning_rate": 7.130041732743014e-05, + "loss": 2.2959, "step": 1630 }, { - "epoch": 0.522337870296237, - "grad_norm": 0.578125, - "learning_rate": 2.741477467574307e-05, - "loss": 1.8618, + "epoch": 0.6336902940676525, + "grad_norm": 0.64453125, + "learning_rate": 7.117044674838527e-05, + "loss": 2.4059, "step": 1631 }, { - "epoch": 0.5226581265012009, - "grad_norm": 0.49609375, - "learning_rate": 2.738694384535046e-05, - "loss": 1.8943, + "epoch": 0.634078822758068, + "grad_norm": 0.64453125, + "learning_rate": 7.104052926213687e-05, + "loss": 2.3168, "step": 1632 }, { - "epoch": 0.5229783827061649, - "grad_norm": 0.73046875, - "learning_rate": 2.73591100293111e-05, - "loss": 2.0352, + "epoch": 0.6344673514484835, + "grad_norm": 0.62890625, + "learning_rate": 7.091066510794224e-05, + "loss": 2.3378, "step": 1633 }, { - "epoch": 0.5232986389111289, - "grad_norm": 0.51953125, - "learning_rate": 2.7331273262440175e-05, - "loss": 2.1791, + "epoch": 0.634855880138899, + "grad_norm": 0.671875, + "learning_rate": 7.07808545249606e-05, + "loss": 2.3458, "step": 1634 }, { - "epoch": 0.5236188951160928, - "grad_norm": 0.9375, - "learning_rate": 2.7303433579556602e-05, - "loss": 2.4775, + "epoch": 0.6352444088293145, + "grad_norm": 0.6328125, + "learning_rate": 7.065109775225231e-05, + "loss": 2.3979, "step": 1635 }, { - "epoch": 0.5239391513210568, - "grad_norm": 0.412109375, - "learning_rate": 2.727559101548292e-05, - "loss": 2.1351, + "epoch": 0.6356329375197299, + "grad_norm": 0.6484375, + "learning_rate": 7.052139502877886e-05, + "loss": 2.3766, "step": 1636 }, { - "epoch": 0.5242594075260208, - "grad_norm": 0.609375, - "learning_rate": 2.7247745605045277e-05, - "loss": 2.3471, + "epoch": 0.6360214662101454, + "grad_norm": 0.65625, + "learning_rate": 7.039174659340202e-05, + "loss": 2.3896, "step": 1637 }, { - "epoch": 0.5245796637309847, - "grad_norm": 1.0390625, - "learning_rate": 2.7219897383073373e-05, - "loss": 1.6935, + "epoch": 0.6364099949005609, + "grad_norm": 0.6796875, + "learning_rate": 7.026215268488367e-05, + "loss": 2.3584, "step": 1638 }, { - "epoch": 0.5248999199359488, - "grad_norm": 0.451171875, - "learning_rate": 2.7192046384400444e-05, - "loss": 2.4096, + "epoch": 0.6367985235909764, + "grad_norm": 0.640625, + "learning_rate": 7.013261354188529e-05, + "loss": 2.3304, "step": 1639 }, { - "epoch": 0.5252201761409128, - "grad_norm": 0.462890625, - "learning_rate": 2.7164192643863196e-05, - "loss": 2.2081, + "epoch": 0.6371870522813919, + "grad_norm": 0.6640625, + "learning_rate": 7.00031294029675e-05, + "loss": 2.3706, "step": 1640 }, { - "epoch": 0.5255404323458767, - "grad_norm": 0.5546875, - "learning_rate": 2.7136336196301737e-05, - "loss": 2.1331, + "epoch": 0.6375755809718073, + "grad_norm": 0.65234375, + "learning_rate": 6.98737005065895e-05, + "loss": 2.3809, "step": 1641 }, { - "epoch": 0.5258606885508407, - "grad_norm": 0.54296875, - "learning_rate": 2.7108477076559595e-05, - "loss": 2.1021, + "epoch": 0.6379641096622228, + "grad_norm": 0.62890625, + "learning_rate": 6.974432709110895e-05, + "loss": 2.337, "step": 1642 }, { - "epoch": 0.5261809447558047, - "grad_norm": 0.546875, - "learning_rate": 2.7080615319483634e-05, - "loss": 2.491, + "epoch": 0.6383526383526383, + "grad_norm": 0.6171875, + "learning_rate": 6.961500939478118e-05, + "loss": 2.4023, "step": 1643 }, { - "epoch": 0.5265012009607686, - "grad_norm": 0.68359375, - "learning_rate": 2.705275095992399e-05, - "loss": 2.7152, + "epoch": 0.6387411670430538, + "grad_norm": 0.67578125, + "learning_rate": 6.948574765575902e-05, + "loss": 2.3571, "step": 1644 }, { - "epoch": 0.5268214571657326, - "grad_norm": 0.5234375, - "learning_rate": 2.7024884032734082e-05, - "loss": 2.138, + "epoch": 0.6391296957334693, + "grad_norm": 0.6328125, + "learning_rate": 6.935654211209213e-05, + "loss": 2.3675, "step": 1645 }, { - "epoch": 0.5271417133706966, - "grad_norm": 0.484375, - "learning_rate": 2.6997014572770534e-05, - "loss": 2.4889, + "epoch": 0.6395182244238848, + "grad_norm": 0.6484375, + "learning_rate": 6.922739300172678e-05, + "loss": 2.3294, "step": 1646 }, { - "epoch": 0.5274619695756605, - "grad_norm": 0.9453125, - "learning_rate": 2.696914261489315e-05, - "loss": 2.4738, + "epoch": 0.6399067531143002, + "grad_norm": 0.65625, + "learning_rate": 6.909830056250527e-05, + "loss": 2.3399, "step": 1647 }, { - "epoch": 0.5277822257806245, - "grad_norm": 0.427734375, - "learning_rate": 2.694126819396482e-05, - "loss": 2.4388, + "epoch": 0.6402952818047157, + "grad_norm": 0.66015625, + "learning_rate": 6.896926503216556e-05, + "loss": 2.4404, "step": 1648 }, { - "epoch": 0.5281024819855885, - "grad_norm": 0.5078125, - "learning_rate": 2.6913391344851556e-05, - "loss": 2.0395, + "epoch": 0.6406838104951312, + "grad_norm": 0.671875, + "learning_rate": 6.884028664834075e-05, + "loss": 2.4031, "step": 1649 }, { - "epoch": 0.5284227381905524, - "grad_norm": 1.125, - "learning_rate": 2.688551210242239e-05, - "loss": 2.4296, + "epoch": 0.6410723391855467, + "grad_norm": 0.66015625, + "learning_rate": 6.871136564855875e-05, + "loss": 2.3244, "step": 1650 }, { - "epoch": 0.5287429943955164, - "grad_norm": 0.453125, - "learning_rate": 2.6857630501549347e-05, - "loss": 2.1873, + "epoch": 0.6414608678759622, + "grad_norm": 0.66796875, + "learning_rate": 6.858250227024178e-05, + "loss": 2.2752, "step": 1651 }, { - "epoch": 0.5290632506004804, - "grad_norm": 0.46484375, - "learning_rate": 2.68297465771074e-05, - "loss": 2.5596, + "epoch": 0.6418493965663777, + "grad_norm": 0.64453125, + "learning_rate": 6.845369675070594e-05, + "loss": 2.3314, "step": 1652 }, { - "epoch": 0.5293835068054443, - "grad_norm": 0.48046875, - "learning_rate": 2.6801860363974434e-05, - "loss": 1.793, + "epoch": 0.6422379252567931, + "grad_norm": 0.62890625, + "learning_rate": 6.832494932716078e-05, + "loss": 2.3836, "step": 1653 }, { - "epoch": 0.5297037630104083, - "grad_norm": 0.58203125, - "learning_rate": 2.6773971897031207e-05, - "loss": 2.0554, + "epoch": 0.6426264539472086, + "grad_norm": 0.64453125, + "learning_rate": 6.819626023670882e-05, + "loss": 2.2433, "step": 1654 }, { - "epoch": 0.5300240192153723, - "grad_norm": 0.37109375, - "learning_rate": 2.6746081211161268e-05, - "loss": 1.7286, + "epoch": 0.6430149826376241, + "grad_norm": 0.64453125, + "learning_rate": 6.806762971634524e-05, + "loss": 2.4247, "step": 1655 }, { - "epoch": 0.5303442754203362, - "grad_norm": 0.396484375, - "learning_rate": 2.6718188341250955e-05, - "loss": 2.0696, + "epoch": 0.6434035113280396, + "grad_norm": 0.66015625, + "learning_rate": 6.793905800295726e-05, + "loss": 2.3075, "step": 1656 }, { - "epoch": 0.5306645316253003, - "grad_norm": 0.51171875, - "learning_rate": 2.6690293322189353e-05, - "loss": 1.9679, + "epoch": 0.6437920400184551, + "grad_norm": 0.6171875, + "learning_rate": 6.78105453333239e-05, + "loss": 2.3406, "step": 1657 }, { - "epoch": 0.5309847878302643, - "grad_norm": 0.5859375, - "learning_rate": 2.6662396188868228e-05, - "loss": 2.3103, + "epoch": 0.6441805687088706, + "grad_norm": 0.66796875, + "learning_rate": 6.768209194411532e-05, + "loss": 2.3682, "step": 1658 }, { - "epoch": 0.5313050440352282, - "grad_norm": 0.51953125, - "learning_rate": 2.6634496976181968e-05, - "loss": 2.3372, + "epoch": 0.644569097399286, + "grad_norm": 0.71875, + "learning_rate": 6.755369807189263e-05, + "loss": 2.3562, "step": 1659 }, { - "epoch": 0.5316253002401922, - "grad_norm": 0.578125, - "learning_rate": 2.6606595719027583e-05, - "loss": 2.3772, + "epoch": 0.6449576260897015, + "grad_norm": 0.70703125, + "learning_rate": 6.742536395310726e-05, + "loss": 2.3898, "step": 1660 }, { - "epoch": 0.5319455564451562, - "grad_norm": 0.734375, - "learning_rate": 2.657869245230464e-05, - "loss": 2.5822, + "epoch": 0.645346154780117, + "grad_norm": 0.625, + "learning_rate": 6.729708982410066e-05, + "loss": 2.4003, "step": 1661 }, { - "epoch": 0.5322658126501201, - "grad_norm": 0.5390625, - "learning_rate": 2.6550787210915223e-05, - "loss": 1.8052, + "epoch": 0.6457346834705325, + "grad_norm": 0.640625, + "learning_rate": 6.71688759211037e-05, + "loss": 2.3614, "step": 1662 }, { - "epoch": 0.5325860688550841, - "grad_norm": 0.69140625, - "learning_rate": 2.6522880029763862e-05, - "loss": 2.583, + "epoch": 0.646123212160948, + "grad_norm": 0.65234375, + "learning_rate": 6.704072248023643e-05, + "loss": 2.3815, "step": 1663 }, { - "epoch": 0.5329063250600481, - "grad_norm": 0.50390625, - "learning_rate": 2.6494970943757548e-05, - "loss": 2.0871, + "epoch": 0.6465117408513635, + "grad_norm": 0.66796875, + "learning_rate": 6.691262973750751e-05, + "loss": 2.3747, "step": 1664 }, { - "epoch": 0.533226581265012, - "grad_norm": 0.5625, - "learning_rate": 2.6467059987805633e-05, - "loss": 2.2921, + "epoch": 0.6469002695417789, + "grad_norm": 0.67578125, + "learning_rate": 6.678459792881387e-05, + "loss": 2.3865, "step": 1665 }, { - "epoch": 0.533546837469976, - "grad_norm": 0.56640625, - "learning_rate": 2.6439147196819795e-05, - "loss": 1.3349, + "epoch": 0.6472887982321944, + "grad_norm": 0.6640625, + "learning_rate": 6.665662728994013e-05, + "loss": 2.2916, "step": 1666 }, { - "epoch": 0.5338670936749399, - "grad_norm": 1.0390625, - "learning_rate": 2.6411232605714043e-05, - "loss": 2.4875, + "epoch": 0.6476773269226099, + "grad_norm": 0.65625, + "learning_rate": 6.652871805655832e-05, + "loss": 2.285, "step": 1667 }, { - "epoch": 0.5341873498799039, - "grad_norm": 0.423828125, - "learning_rate": 2.6383316249404615e-05, - "loss": 1.9192, + "epoch": 0.6480658556130254, + "grad_norm": 0.62890625, + "learning_rate": 6.640087046422738e-05, + "loss": 2.3718, "step": 1668 }, { - "epoch": 0.5345076060848679, - "grad_norm": 0.427734375, - "learning_rate": 2.6355398162809957e-05, - "loss": 2.4075, + "epoch": 0.6484543843034409, + "grad_norm": 0.66796875, + "learning_rate": 6.627308474839277e-05, + "loss": 2.3982, "step": 1669 }, { - "epoch": 0.5348278622898318, - "grad_norm": 0.77734375, - "learning_rate": 2.632747838085068e-05, - "loss": 2.5198, + "epoch": 0.6488429129938564, + "grad_norm": 0.65234375, + "learning_rate": 6.614536114438589e-05, + "loss": 2.4418, "step": 1670 }, { - "epoch": 0.5351481184947958, - "grad_norm": 0.423828125, - "learning_rate": 2.6299556938449528e-05, - "loss": 2.4702, + "epoch": 0.6492314416842718, + "grad_norm": 0.66796875, + "learning_rate": 6.601769988742386e-05, + "loss": 2.3537, "step": 1671 }, { - "epoch": 0.5354683746997598, - "grad_norm": 0.56640625, - "learning_rate": 2.627163387053131e-05, - "loss": 2.4223, + "epoch": 0.6496199703746873, + "grad_norm": 0.6796875, + "learning_rate": 6.58901012126089e-05, + "loss": 2.3693, "step": 1672 }, { - "epoch": 0.5357886309047237, - "grad_norm": 0.546875, - "learning_rate": 2.624370921202286e-05, - "loss": 2.0182, + "epoch": 0.6500084990651028, + "grad_norm": 0.640625, + "learning_rate": 6.576256535492806e-05, + "loss": 2.3887, "step": 1673 }, { - "epoch": 0.5361088871096877, - "grad_norm": 0.6328125, - "learning_rate": 2.6215782997853026e-05, - "loss": 2.2894, + "epoch": 0.6503970277555183, + "grad_norm": 0.67578125, + "learning_rate": 6.563509254925268e-05, + "loss": 2.3907, "step": 1674 }, { - "epoch": 0.5364291433146517, - "grad_norm": 0.58203125, - "learning_rate": 2.6187855262952587e-05, - "loss": 2.3239, + "epoch": 0.6507855564459338, + "grad_norm": 0.65234375, + "learning_rate": 6.550768303033789e-05, + "loss": 2.398, "step": 1675 }, { - "epoch": 0.5367493995196156, - "grad_norm": 0.78125, - "learning_rate": 2.615992604225422e-05, - "loss": 2.5619, + "epoch": 0.6511740851363493, + "grad_norm": 0.71875, + "learning_rate": 6.538033703282243e-05, + "loss": 2.3144, "step": 1676 }, { - "epoch": 0.5370696557245797, - "grad_norm": 0.6328125, - "learning_rate": 2.6131995370692463e-05, - "loss": 1.886, + "epoch": 0.6515626138267647, + "grad_norm": 0.6640625, + "learning_rate": 6.52530547912279e-05, + "loss": 2.3954, "step": 1677 }, { - "epoch": 0.5373899119295437, - "grad_norm": 0.8125, - "learning_rate": 2.6104063283203672e-05, - "loss": 2.2598, + "epoch": 0.6519511425171802, + "grad_norm": 0.68359375, + "learning_rate": 6.512583653995867e-05, + "loss": 2.4037, "step": 1678 }, { - "epoch": 0.5377101681345076, - "grad_norm": 0.408203125, - "learning_rate": 2.607612981472599e-05, - "loss": 1.7341, + "epoch": 0.6523396712075957, + "grad_norm": 0.73828125, + "learning_rate": 6.499868251330102e-05, + "loss": 2.3779, "step": 1679 }, { - "epoch": 0.5380304243394716, - "grad_norm": 1.328125, - "learning_rate": 2.6048195000199248e-05, - "loss": 1.8696, + "epoch": 0.6527281998980112, + "grad_norm": 0.65234375, + "learning_rate": 6.48715929454232e-05, + "loss": 2.3817, "step": 1680 }, { - "epoch": 0.5383506805444356, - "grad_norm": 0.51953125, - "learning_rate": 2.6020258874565002e-05, - "loss": 2.6784, + "epoch": 0.6531167285884267, + "grad_norm": 0.66796875, + "learning_rate": 6.474456807037456e-05, + "loss": 2.3627, "step": 1681 }, { - "epoch": 0.5386709367493995, - "grad_norm": 0.765625, - "learning_rate": 2.5992321472766418e-05, - "loss": 1.9824, + "epoch": 0.6535052572788422, + "grad_norm": 0.6640625, + "learning_rate": 6.461760812208544e-05, + "loss": 2.3574, "step": 1682 }, { - "epoch": 0.5389911929543635, - "grad_norm": 0.56640625, - "learning_rate": 2.596438282974828e-05, - "loss": 2.1182, + "epoch": 0.6538937859692576, + "grad_norm": 0.6796875, + "learning_rate": 6.449071333436654e-05, + "loss": 2.3536, "step": 1683 }, { - "epoch": 0.5393114491593275, - "grad_norm": 0.51953125, - "learning_rate": 2.59364429804569e-05, - "loss": 1.9854, + "epoch": 0.6542823146596731, + "grad_norm": 0.6640625, + "learning_rate": 6.436388394090862e-05, + "loss": 2.3958, "step": 1684 }, { - "epoch": 0.5396317053642914, - "grad_norm": 0.46484375, - "learning_rate": 2.5908501959840125e-05, - "loss": 2.27, + "epoch": 0.6546708433500886, + "grad_norm": 0.640625, + "learning_rate": 6.423712017528193e-05, + "loss": 2.3059, "step": 1685 }, { - "epoch": 0.5399519615692554, - "grad_norm": 0.6796875, - "learning_rate": 2.588055980284725e-05, - "loss": 2.0739, + "epoch": 0.6550593720405041, + "grad_norm": 0.66015625, + "learning_rate": 6.411042227093592e-05, + "loss": 2.4365, "step": 1686 }, { - "epoch": 0.5402722177742194, - "grad_norm": 0.50390625, - "learning_rate": 2.5852616544429008e-05, - "loss": 2.2536, + "epoch": 0.6554479007309196, + "grad_norm": 0.67578125, + "learning_rate": 6.398379046119873e-05, + "loss": 2.3754, "step": 1687 }, { - "epoch": 0.5405924739791833, - "grad_norm": 0.5234375, - "learning_rate": 2.5824672219537483e-05, - "loss": 2.2221, + "epoch": 0.6558364294213351, + "grad_norm": 0.65625, + "learning_rate": 6.38572249792768e-05, + "loss": 2.3531, "step": 1688 }, { - "epoch": 0.5409127301841473, - "grad_norm": 0.72265625, - "learning_rate": 2.5796726863126113e-05, - "loss": 2.7278, + "epoch": 0.6562249581117505, + "grad_norm": 0.6171875, + "learning_rate": 6.373072605825435e-05, + "loss": 2.3918, "step": 1689 }, { - "epoch": 0.5412329863891113, - "grad_norm": 0.5703125, - "learning_rate": 2.5768780510149633e-05, - "loss": 2.2353, + "epoch": 0.656613486802166, + "grad_norm": 0.62109375, + "learning_rate": 6.360429393109311e-05, + "loss": 2.2982, "step": 1690 }, { - "epoch": 0.5415532425940752, - "grad_norm": 0.4140625, - "learning_rate": 2.5740833195563996e-05, - "loss": 1.9356, + "epoch": 0.6570020154925815, + "grad_norm": 0.70703125, + "learning_rate": 6.347792883063173e-05, + "loss": 2.4119, "step": 1691 }, { - "epoch": 0.5418734987990392, - "grad_norm": 0.52734375, - "learning_rate": 2.571288495432639e-05, - "loss": 1.7192, + "epoch": 0.657390544182997, + "grad_norm": 0.65625, + "learning_rate": 6.335163098958552e-05, + "loss": 2.3786, "step": 1692 }, { - "epoch": 0.5421937550040032, - "grad_norm": 0.50390625, - "learning_rate": 2.5684935821395135e-05, - "loss": 2.4505, + "epoch": 0.6577790728734125, + "grad_norm": 0.703125, + "learning_rate": 6.322540064054578e-05, + "loss": 2.3941, "step": 1693 }, { - "epoch": 0.5425140112089671, - "grad_norm": 0.427734375, - "learning_rate": 2.5656985831729692e-05, - "loss": 1.8894, + "epoch": 0.658167601563828, + "grad_norm": 0.68359375, + "learning_rate": 6.309923801597965e-05, + "loss": 2.4113, "step": 1694 }, { - "epoch": 0.5428342674139311, - "grad_norm": 0.4296875, - "learning_rate": 2.5629035020290566e-05, - "loss": 1.8297, + "epoch": 0.6585561302542434, + "grad_norm": 0.6328125, + "learning_rate": 6.297314334822945e-05, + "loss": 2.3399, "step": 1695 }, { - "epoch": 0.5431545236188952, - "grad_norm": 0.43359375, - "learning_rate": 2.5601083422039314e-05, - "loss": 2.2753, + "epoch": 0.6589446589446589, + "grad_norm": 0.62890625, + "learning_rate": 6.284711686951243e-05, + "loss": 2.385, "step": 1696 }, { - "epoch": 0.543474779823859, - "grad_norm": 0.51171875, - "learning_rate": 2.5573131071938456e-05, - "loss": 2.8383, + "epoch": 0.6593331876350744, + "grad_norm": 0.72265625, + "learning_rate": 6.272115881192025e-05, + "loss": 2.3508, "step": 1697 }, { - "epoch": 0.5437950360288231, - "grad_norm": 0.44921875, - "learning_rate": 2.554517800495148e-05, - "loss": 2.4237, + "epoch": 0.6597217163254899, + "grad_norm": 0.74609375, + "learning_rate": 6.259526940741847e-05, + "loss": 2.3601, "step": 1698 }, { - "epoch": 0.5441152922337871, - "grad_norm": 0.40625, - "learning_rate": 2.5517224256042745e-05, - "loss": 2.0334, + "epoch": 0.6601102450159054, + "grad_norm": 0.734375, + "learning_rate": 6.246944888784634e-05, + "loss": 2.3568, "step": 1699 }, { - "epoch": 0.544435548438751, - "grad_norm": 0.5078125, - "learning_rate": 2.5489269860177474e-05, - "loss": 2.4893, + "epoch": 0.6604987737063209, + "grad_norm": 0.63671875, + "learning_rate": 6.234369748491617e-05, + "loss": 2.3455, "step": 1700 }, { - "epoch": 0.544755804643715, + "epoch": 0.6608873023967363, "grad_norm": 0.60546875, - "learning_rate": 2.5461314852321694e-05, - "loss": 2.0231, + "learning_rate": 6.221801543021306e-05, + "loss": 2.2836, "step": 1701 }, { - "epoch": 0.545076060848679, - "grad_norm": 0.5390625, - "learning_rate": 2.5433359267442204e-05, - "loss": 2.486, + "epoch": 0.6612758310871518, + "grad_norm": 0.671875, + "learning_rate": 6.209240295519428e-05, + "loss": 2.4369, "step": 1702 }, { - "epoch": 0.5453963170536429, - "grad_norm": 0.79296875, - "learning_rate": 2.5405403140506517e-05, - "loss": 2.3233, + "epoch": 0.6616643597775673, + "grad_norm": 0.734375, + "learning_rate": 6.196686029118909e-05, + "loss": 2.3962, "step": 1703 }, { - "epoch": 0.5457165732586069, - "grad_norm": 0.53515625, - "learning_rate": 2.5377446506482837e-05, - "loss": 2.0646, + "epoch": 0.6620528884679828, + "grad_norm": 0.6640625, + "learning_rate": 6.184138766939811e-05, + "loss": 2.3827, "step": 1704 }, { - "epoch": 0.5460368294635709, - "grad_norm": 0.439453125, - "learning_rate": 2.5349489400339997e-05, - "loss": 2.1051, + "epoch": 0.6624414171583983, + "grad_norm": 0.66015625, + "learning_rate": 6.171598532089299e-05, + "loss": 2.2286, "step": 1705 }, { - "epoch": 0.5463570856685348, - "grad_norm": 0.5078125, - "learning_rate": 2.5321531857047404e-05, - "loss": 2.5043, + "epoch": 0.6628299458488138, + "grad_norm": 0.6875, + "learning_rate": 6.159065347661596e-05, + "loss": 2.3438, "step": 1706 }, { - "epoch": 0.5466773418734988, - "grad_norm": 0.494140625, - "learning_rate": 2.5293573911575036e-05, - "loss": 1.8195, + "epoch": 0.6632184745392292, + "grad_norm": 0.6484375, + "learning_rate": 6.14653923673794e-05, + "loss": 2.4297, "step": 1707 }, { - "epoch": 0.5469975980784628, - "grad_norm": 0.50390625, - "learning_rate": 2.5265615598893356e-05, - "loss": 2.6448, + "epoch": 0.6636070032296447, + "grad_norm": 0.61328125, + "learning_rate": 6.134020222386544e-05, + "loss": 2.3676, "step": 1708 }, { - "epoch": 0.5473178542834267, - "grad_norm": 0.53515625, - "learning_rate": 2.5237656953973303e-05, - "loss": 2.3232, + "epoch": 0.6639955319200602, + "grad_norm": 0.65625, + "learning_rate": 6.121508327662553e-05, + "loss": 2.3893, "step": 1709 }, { - "epoch": 0.5476381104883907, - "grad_norm": 0.63671875, - "learning_rate": 2.5209698011786214e-05, - "loss": 1.9998, + "epoch": 0.6643840606104757, + "grad_norm": 0.6640625, + "learning_rate": 6.109003575607997e-05, + "loss": 2.4231, "step": 1710 }, { - "epoch": 0.5479583666933547, - "grad_norm": 0.5859375, - "learning_rate": 2.5181738807303806e-05, - "loss": 2.0763, + "epoch": 0.6647725893008912, + "grad_norm": 0.65234375, + "learning_rate": 6.096505989251753e-05, + "loss": 2.3627, "step": 1711 }, { - "epoch": 0.5482786228983186, - "grad_norm": 0.58203125, - "learning_rate": 2.5153779375498137e-05, - "loss": 2.3951, + "epoch": 0.6651611179913066, + "grad_norm": 0.6484375, + "learning_rate": 6.0840155916095044e-05, + "loss": 2.3278, "step": 1712 }, { - "epoch": 0.5485988791032826, - "grad_norm": 0.71484375, - "learning_rate": 2.5125819751341522e-05, - "loss": 1.9252, + "epoch": 0.6655496466817221, + "grad_norm": 0.61328125, + "learning_rate": 6.071532405683691e-05, + "loss": 2.3325, "step": 1713 }, { - "epoch": 0.5489191353082465, - "grad_norm": 0.546875, - "learning_rate": 2.509785996980653e-05, - "loss": 2.0553, + "epoch": 0.6659381753721376, + "grad_norm": 0.6640625, + "learning_rate": 6.059056454463479e-05, + "loss": 2.3673, "step": 1714 }, { - "epoch": 0.5492393915132106, - "grad_norm": 0.490234375, - "learning_rate": 2.506990006586594e-05, - "loss": 2.1069, + "epoch": 0.6663267040625531, + "grad_norm": 0.625, + "learning_rate": 6.046587760924698e-05, + "loss": 2.308, "step": 1715 }, { - "epoch": 0.5495596477181746, - "grad_norm": 0.6171875, - "learning_rate": 2.5041940074492676e-05, - "loss": 2.3812, + "epoch": 0.6667152327529686, + "grad_norm": 0.6328125, + "learning_rate": 6.034126348029827e-05, + "loss": 2.3354, "step": 1716 }, { - "epoch": 0.5498799039231385, - "grad_norm": 0.5078125, - "learning_rate": 2.5013980030659743e-05, - "loss": 2.1561, + "epoch": 0.6671037614433841, + "grad_norm": 0.62890625, + "learning_rate": 6.021672238727927e-05, + "loss": 2.3802, "step": 1717 }, { - "epoch": 0.5502001601281025, - "grad_norm": 0.51953125, - "learning_rate": 2.498601996934026e-05, - "loss": 1.9662, + "epoch": 0.6674922901337995, + "grad_norm": 0.671875, + "learning_rate": 6.009225455954614e-05, + "loss": 2.3173, "step": 1718 }, { - "epoch": 0.5505204163330665, - "grad_norm": 0.421875, - "learning_rate": 2.4958059925507333e-05, - "loss": 1.932, + "epoch": 0.667880818824215, + "grad_norm": 0.6484375, + "learning_rate": 5.996786022632004e-05, + "loss": 2.417, "step": 1719 }, { - "epoch": 0.5508406725380304, - "grad_norm": 0.85546875, - "learning_rate": 2.4930099934134065e-05, - "loss": 2.3476, + "epoch": 0.6682693475146305, + "grad_norm": 0.65234375, + "learning_rate": 5.9843539616686875e-05, + "loss": 2.3546, "step": 1720 }, { - "epoch": 0.5511609287429944, - "grad_norm": 0.5078125, - "learning_rate": 2.490214003019347e-05, - "loss": 2.2002, + "epoch": 0.668657876205046, + "grad_norm": 0.67578125, + "learning_rate": 5.971929295959671e-05, + "loss": 2.3215, "step": 1721 }, { - "epoch": 0.5514811849479584, - "grad_norm": 0.51171875, - "learning_rate": 2.4874180248658484e-05, - "loss": 2.3532, + "epoch": 0.6690464048954615, + "grad_norm": 0.6328125, + "learning_rate": 5.959512048386349e-05, + "loss": 2.3583, "step": 1722 }, { - "epoch": 0.5518014411529223, - "grad_norm": 0.53125, - "learning_rate": 2.4846220624501872e-05, - "loss": 2.258, + "epoch": 0.669434933585877, + "grad_norm": 0.65625, + "learning_rate": 5.9471022418164425e-05, + "loss": 2.4115, "step": 1723 }, { - "epoch": 0.5521216973578863, - "grad_norm": 0.61328125, - "learning_rate": 2.48182611926962e-05, - "loss": 2.4362, + "epoch": 0.6698234622762924, + "grad_norm": 0.65234375, + "learning_rate": 5.934699899103982e-05, + "loss": 2.4339, "step": 1724 }, { - "epoch": 0.5524419535628503, - "grad_norm": 0.6171875, - "learning_rate": 2.479030198821379e-05, - "loss": 1.7804, + "epoch": 0.6702119909667079, + "grad_norm": 0.63671875, + "learning_rate": 5.9223050430892445e-05, + "loss": 2.3634, "step": 1725 }, { - "epoch": 0.5527622097678142, - "grad_norm": 0.515625, - "learning_rate": 2.4762343046026702e-05, - "loss": 2.7527, + "epoch": 0.6706005196571234, + "grad_norm": 0.69921875, + "learning_rate": 5.909917696598721e-05, + "loss": 2.373, "step": 1726 }, { - "epoch": 0.5530824659727782, - "grad_norm": 0.625, - "learning_rate": 2.473438440110665e-05, - "loss": 2.4608, + "epoch": 0.6709890483475389, + "grad_norm": 0.671875, + "learning_rate": 5.897537882445079e-05, + "loss": 2.4134, "step": 1727 }, { - "epoch": 0.5534027221777422, - "grad_norm": 0.380859375, - "learning_rate": 2.4706426088424973e-05, - "loss": 1.8141, + "epoch": 0.6713775770379544, + "grad_norm": 0.62109375, + "learning_rate": 5.8851656234271055e-05, + "loss": 2.3665, "step": 1728 }, { - "epoch": 0.5537229783827061, - "grad_norm": 0.53515625, - "learning_rate": 2.46784681429526e-05, - "loss": 2.2575, + "epoch": 0.6717661057283699, + "grad_norm": 0.6484375, + "learning_rate": 5.872800942329677e-05, + "loss": 2.3632, "step": 1729 }, { - "epoch": 0.5540432345876701, - "grad_norm": 0.33203125, - "learning_rate": 2.4650510599660005e-05, - "loss": 2.6203, + "epoch": 0.6721546344187853, + "grad_norm": 0.67578125, + "learning_rate": 5.860443861923718e-05, + "loss": 2.3408, "step": 1730 }, { - "epoch": 0.5543634907926341, - "grad_norm": 0.5234375, - "learning_rate": 2.4622553493517165e-05, - "loss": 2.1001, + "epoch": 0.6725431631092008, + "grad_norm": 0.625, + "learning_rate": 5.848094404966149e-05, + "loss": 2.3687, "step": 1731 }, { - "epoch": 0.554683746997598, - "grad_norm": 0.6796875, - "learning_rate": 2.4594596859493485e-05, - "loss": 2.3472, + "epoch": 0.6729316917996163, + "grad_norm": 0.69140625, + "learning_rate": 5.83575259419986e-05, + "loss": 2.4013, "step": 1732 }, { - "epoch": 0.555004003202562, - "grad_norm": 0.462890625, - "learning_rate": 2.4566640732557802e-05, - "loss": 2.0663, + "epoch": 0.6733202204900318, + "grad_norm": 0.671875, + "learning_rate": 5.82341845235365e-05, + "loss": 2.3826, "step": 1733 }, { - "epoch": 0.555324259407526, - "grad_norm": 0.53515625, - "learning_rate": 2.4538685147678316e-05, - "loss": 2.3392, + "epoch": 0.6737087491804473, + "grad_norm": 0.64453125, + "learning_rate": 5.811092002142198e-05, + "loss": 2.4015, "step": 1734 }, { - "epoch": 0.55564451561249, - "grad_norm": 0.70703125, - "learning_rate": 2.4510730139822535e-05, - "loss": 2.1949, + "epoch": 0.6740972778708628, + "grad_norm": 0.62890625, + "learning_rate": 5.798773266266022e-05, + "loss": 2.3255, "step": 1735 }, { - "epoch": 0.555964771817454, - "grad_norm": 0.478515625, - "learning_rate": 2.4482775743957258e-05, - "loss": 2.517, + "epoch": 0.6744858065612782, + "grad_norm": 0.625, + "learning_rate": 5.786462267411431e-05, + "loss": 2.3693, "step": 1736 }, { - "epoch": 0.556285028022418, - "grad_norm": 0.43359375, - "learning_rate": 2.4454821995048522e-05, - "loss": 2.4129, + "epoch": 0.6748743352516937, + "grad_norm": 0.6484375, + "learning_rate": 5.774159028250483e-05, + "loss": 2.407, "step": 1737 }, { - "epoch": 0.5566052842273819, - "grad_norm": 0.7109375, - "learning_rate": 2.442686892806155e-05, - "loss": 2.5147, + "epoch": 0.6752628639421092, + "grad_norm": 0.71875, + "learning_rate": 5.761863571440949e-05, + "loss": 2.4144, "step": 1738 }, { - "epoch": 0.5569255404323459, - "grad_norm": 0.578125, - "learning_rate": 2.4398916577960695e-05, - "loss": 2.8204, + "epoch": 0.6756513926325247, + "grad_norm": 0.66015625, + "learning_rate": 5.749575919626269e-05, + "loss": 2.3344, "step": 1739 }, { - "epoch": 0.5572457966373099, - "grad_norm": 0.5859375, - "learning_rate": 2.437096497970944e-05, - "loss": 2.6784, + "epoch": 0.6760399213229402, + "grad_norm": 0.65625, + "learning_rate": 5.737296095435498e-05, + "loss": 2.3339, "step": 1740 }, { - "epoch": 0.5575660528422738, - "grad_norm": 0.453125, - "learning_rate": 2.4343014168270314e-05, - "loss": 2.1789, + "epoch": 0.6764284500133557, + "grad_norm": 0.7109375, + "learning_rate": 5.7250241214832914e-05, + "loss": 2.3194, "step": 1741 }, { - "epoch": 0.5578863090472378, - "grad_norm": 0.466796875, - "learning_rate": 2.431506417860487e-05, - "loss": 2.8155, + "epoch": 0.6768169787037711, + "grad_norm": 0.6796875, + "learning_rate": 5.7127600203698385e-05, + "loss": 2.3024, "step": 1742 }, { - "epoch": 0.5582065652522018, - "grad_norm": 0.4765625, - "learning_rate": 2.4287115045673617e-05, - "loss": 2.0444, + "epoch": 0.6772055073941866, + "grad_norm": 0.6796875, + "learning_rate": 5.700503814680831e-05, + "loss": 2.3236, "step": 1743 }, { - "epoch": 0.5585268214571657, - "grad_norm": 0.59375, - "learning_rate": 2.4259166804436006e-05, - "loss": 2.646, + "epoch": 0.6775940360846021, + "grad_norm": 0.609375, + "learning_rate": 5.6882555269874224e-05, + "loss": 2.3919, "step": 1744 }, { - "epoch": 0.5588470776621297, - "grad_norm": 0.5859375, - "learning_rate": 2.4231219489850376e-05, - "loss": 2.2975, + "epoch": 0.6779825647750176, + "grad_norm": 0.6328125, + "learning_rate": 5.676015179846181e-05, + "loss": 2.3636, "step": 1745 }, { - "epoch": 0.5591673338670937, - "grad_norm": 0.765625, - "learning_rate": 2.4203273136873892e-05, - "loss": 2.1118, + "epoch": 0.6783710934654331, + "grad_norm": 0.6484375, + "learning_rate": 5.663782795799057e-05, + "loss": 2.397, "step": 1746 }, { - "epoch": 0.5594875900720576, - "grad_norm": 0.462890625, - "learning_rate": 2.4175327780462523e-05, - "loss": 2.1737, + "epoch": 0.6787596221558486, + "grad_norm": 0.6953125, + "learning_rate": 5.651558397373335e-05, + "loss": 2.3954, "step": 1747 }, { - "epoch": 0.5598078462770216, - "grad_norm": 0.412109375, - "learning_rate": 2.4147383455571004e-05, - "loss": 2.2012, + "epoch": 0.679148150846264, + "grad_norm": 0.62890625, + "learning_rate": 5.6393420070815806e-05, + "loss": 2.3431, "step": 1748 }, { - "epoch": 0.5601281024819856, - "grad_norm": 0.5625, - "learning_rate": 2.4119440197152754e-05, - "loss": 1.4749, + "epoch": 0.6795366795366795, + "grad_norm": 0.65625, + "learning_rate": 5.627133647421628e-05, + "loss": 2.3616, "step": 1749 }, { - "epoch": 0.5604483586869495, + "epoch": 0.679925208227095, "grad_norm": 0.640625, - "learning_rate": 2.409149804015988e-05, - "loss": 2.1635, + "learning_rate": 5.614933340876515e-05, + "loss": 2.4101, "step": 1750 }, { - "epoch": 0.5607686148919135, - "grad_norm": 0.39453125, - "learning_rate": 2.4063557019543107e-05, - "loss": 1.4912, + "epoch": 0.6803137369175105, + "grad_norm": 0.64453125, + "learning_rate": 5.602741109914449e-05, + "loss": 2.3575, "step": 1751 }, { - "epoch": 0.5610888710968776, - "grad_norm": 0.5078125, - "learning_rate": 2.403561717025173e-05, - "loss": 2.4878, + "epoch": 0.680702265607926, + "grad_norm": 0.6640625, + "learning_rate": 5.590556976988766e-05, + "loss": 2.396, "step": 1752 }, { - "epoch": 0.5614091273018414, - "grad_norm": 0.55078125, - "learning_rate": 2.400767852723359e-05, - "loss": 2.3552, + "epoch": 0.6810907942983415, + "grad_norm": 0.671875, + "learning_rate": 5.578380964537888e-05, + "loss": 2.3252, "step": 1753 }, { - "epoch": 0.5617293835068055, - "grad_norm": 0.4765625, - "learning_rate": 2.3979741125435004e-05, - "loss": 1.6363, + "epoch": 0.6814793229887569, + "grad_norm": 0.609375, + "learning_rate": 5.566213094985282e-05, + "loss": 2.3266, "step": 1754 }, { - "epoch": 0.5620496397117695, - "grad_norm": 1.078125, - "learning_rate": 2.3951804999800754e-05, - "loss": 2.0071, + "epoch": 0.6818678516791724, + "grad_norm": 0.6484375, + "learning_rate": 5.55405339073942e-05, + "loss": 2.3403, "step": 1755 }, { - "epoch": 0.5623698959167334, - "grad_norm": 0.5625, - "learning_rate": 2.3923870185274018e-05, - "loss": 2.3637, + "epoch": 0.6822563803695879, + "grad_norm": 0.6484375, + "learning_rate": 5.5419018741937424e-05, + "loss": 2.3728, "step": 1756 }, { - "epoch": 0.5626901521216974, - "grad_norm": 0.494140625, - "learning_rate": 2.389593671679633e-05, - "loss": 2.4539, + "epoch": 0.6826449090600034, + "grad_norm": 0.67578125, + "learning_rate": 5.529758567726593e-05, + "loss": 2.3492, "step": 1757 }, { - "epoch": 0.5630104083266614, - "grad_norm": 0.271484375, - "learning_rate": 2.3868004629307543e-05, - "loss": 2.087, + "epoch": 0.6830334377504189, + "grad_norm": 0.6953125, + "learning_rate": 5.517623493701213e-05, + "loss": 2.3451, "step": 1758 }, { - "epoch": 0.5633306645316253, - "grad_norm": 0.83984375, - "learning_rate": 2.3840073957745793e-05, - "loss": 2.3264, + "epoch": 0.6834219664408344, + "grad_norm": 0.6328125, + "learning_rate": 5.505496674465678e-05, + "loss": 2.3361, "step": 1759 }, { - "epoch": 0.5636509207365893, - "grad_norm": 0.55859375, - "learning_rate": 2.381214473704742e-05, - "loss": 2.3831, + "epoch": 0.6838104951312498, + "grad_norm": 0.62890625, + "learning_rate": 5.49337813235286e-05, + "loss": 2.3548, "step": 1760 }, { - "epoch": 0.5639711769415532, - "grad_norm": 0.52734375, - "learning_rate": 2.3784217002146976e-05, - "loss": 2.5073, + "epoch": 0.6841990238216653, + "grad_norm": 0.63671875, + "learning_rate": 5.481267889680388e-05, + "loss": 2.3969, "step": 1761 }, { - "epoch": 0.5642914331465172, - "grad_norm": 0.494140625, - "learning_rate": 2.3756290787977147e-05, - "loss": 2.3088, + "epoch": 0.6845875525120808, + "grad_norm": 0.64453125, + "learning_rate": 5.4691659687506094e-05, + "loss": 2.3452, "step": 1762 }, { - "epoch": 0.5646116893514812, + "epoch": 0.6849760812024963, "grad_norm": 0.69921875, - "learning_rate": 2.3728366129468696e-05, - "loss": 2.3283, + "learning_rate": 5.457072391850543e-05, + "loss": 2.3758, "step": 1763 }, { - "epoch": 0.5649319455564451, - "grad_norm": 0.46484375, - "learning_rate": 2.3700443061550478e-05, - "loss": 2.4001, + "epoch": 0.6853646098929118, + "grad_norm": 0.65234375, + "learning_rate": 5.444987181251845e-05, + "loss": 2.4076, "step": 1764 }, { - "epoch": 0.5652522017614091, - "grad_norm": 0.50390625, - "learning_rate": 2.3672521619149322e-05, - "loss": 2.3446, + "epoch": 0.6857531385833273, + "grad_norm": 0.66796875, + "learning_rate": 5.432910359210754e-05, + "loss": 2.3528, "step": 1765 }, { - "epoch": 0.5655724579663731, - "grad_norm": 0.58203125, - "learning_rate": 2.364460183719005e-05, - "loss": 2.1358, + "epoch": 0.6861416672737427, + "grad_norm": 0.62890625, + "learning_rate": 5.420841947968076e-05, + "loss": 2.3899, "step": 1766 }, { - "epoch": 0.565892714171337, - "grad_norm": 0.57421875, - "learning_rate": 2.3616683750595394e-05, - "loss": 2.2924, + "epoch": 0.6865301959641582, + "grad_norm": 0.62109375, + "learning_rate": 5.4087819697491116e-05, + "loss": 2.3693, "step": 1767 }, { - "epoch": 0.566212970376301, - "grad_norm": 0.494140625, - "learning_rate": 2.3588767394285963e-05, - "loss": 1.6567, + "epoch": 0.6869187246545737, + "grad_norm": 0.6484375, + "learning_rate": 5.3967304467636406e-05, + "loss": 2.3826, "step": 1768 }, { - "epoch": 0.566533226581265, - "grad_norm": 0.462890625, - "learning_rate": 2.356085280318021e-05, - "loss": 2.1492, + "epoch": 0.6873072533449892, + "grad_norm": 0.640625, + "learning_rate": 5.38468740120587e-05, + "loss": 2.3899, "step": 1769 }, { - "epoch": 0.5668534827862289, - "grad_norm": 0.42578125, - "learning_rate": 2.3532940012194383e-05, - "loss": 2.6153, + "epoch": 0.6876957820354047, + "grad_norm": 0.609375, + "learning_rate": 5.372652855254394e-05, + "loss": 2.3605, "step": 1770 }, { - "epoch": 0.5671737389911929, - "grad_norm": 0.490234375, - "learning_rate": 2.3505029056242454e-05, - "loss": 2.1282, + "epoch": 0.6880843107258202, + "grad_norm": 0.640625, + "learning_rate": 5.360626831072156e-05, + "loss": 2.4103, "step": 1771 }, { - "epoch": 0.567493995196157, - "grad_norm": 0.5546875, - "learning_rate": 2.347711997023614e-05, - "loss": 2.281, + "epoch": 0.6884728394162356, + "grad_norm": 0.671875, + "learning_rate": 5.348609350806402e-05, + "loss": 2.3197, "step": 1772 }, { - "epoch": 0.5678142514011209, - "grad_norm": 0.5859375, - "learning_rate": 2.344921278908479e-05, - "loss": 2.5034, + "epoch": 0.6888613681066511, + "grad_norm": 0.62109375, + "learning_rate": 5.3366004365886433e-05, + "loss": 2.3234, "step": 1773 }, { - "epoch": 0.5681345076060849, - "grad_norm": 0.6640625, - "learning_rate": 2.342130754769536e-05, - "loss": 2.2516, + "epoch": 0.6892498967970666, + "grad_norm": 0.66015625, + "learning_rate": 5.3246001105346254e-05, + "loss": 2.3023, "step": 1774 }, { - "epoch": 0.5684547638110489, - "grad_norm": 0.484375, - "learning_rate": 2.3393404280972426e-05, - "loss": 2.1612, + "epoch": 0.6896384254874821, + "grad_norm": 0.68359375, + "learning_rate": 5.3126083947442584e-05, + "loss": 2.4167, "step": 1775 }, { - "epoch": 0.5687750200160128, - "grad_norm": 0.470703125, - "learning_rate": 2.336550302381804e-05, - "loss": 2.4927, + "epoch": 0.6900269541778976, + "grad_norm": 0.61328125, + "learning_rate": 5.300625311301615e-05, + "loss": 2.3408, "step": 1776 }, { - "epoch": 0.5690952762209768, - "grad_norm": 0.47265625, - "learning_rate": 2.333760381113178e-05, - "loss": 2.2154, + "epoch": 0.690415482868313, + "grad_norm": 0.62890625, + "learning_rate": 5.2886508822748585e-05, + "loss": 2.347, "step": 1777 }, { - "epoch": 0.5694155324259408, - "grad_norm": 0.48046875, - "learning_rate": 2.330970667781065e-05, - "loss": 1.8648, + "epoch": 0.6908040115587285, + "grad_norm": 0.60546875, + "learning_rate": 5.27668512971622e-05, + "loss": 2.3668, "step": 1778 }, { - "epoch": 0.5697357886309047, - "grad_norm": 0.5859375, - "learning_rate": 2.328181165874905e-05, - "loss": 2.3315, + "epoch": 0.691192540249144, + "grad_norm": 0.625, + "learning_rate": 5.264728075661951e-05, + "loss": 2.3325, "step": 1779 }, { - "epoch": 0.5700560448358687, - "grad_norm": 0.46484375, - "learning_rate": 2.325391878883874e-05, - "loss": 2.1458, + "epoch": 0.6915810689395595, + "grad_norm": 0.625, + "learning_rate": 5.2527797421322835e-05, + "loss": 2.3634, "step": 1780 }, { - "epoch": 0.5703763010408327, - "grad_norm": 0.439453125, - "learning_rate": 2.3226028102968806e-05, - "loss": 1.9153, + "epoch": 0.691969597629975, + "grad_norm": 0.6953125, + "learning_rate": 5.240840151131388e-05, + "loss": 2.3602, "step": 1781 }, { - "epoch": 0.5706965572457966, - "grad_norm": 0.59375, - "learning_rate": 2.3198139636025565e-05, - "loss": 1.8971, + "epoch": 0.6923581263203905, + "grad_norm": 0.64453125, + "learning_rate": 5.2289093246473355e-05, + "loss": 2.425, "step": 1782 }, { - "epoch": 0.5710168134507606, - "grad_norm": 0.53125, - "learning_rate": 2.3170253422892606e-05, - "loss": 2.1762, + "epoch": 0.6927466550108059, + "grad_norm": 0.6328125, + "learning_rate": 5.216987284652061e-05, + "loss": 2.3902, "step": 1783 }, { - "epoch": 0.5713370696557246, - "grad_norm": 0.43359375, - "learning_rate": 2.3142369498450665e-05, - "loss": 1.8368, + "epoch": 0.6931351837012214, + "grad_norm": 0.66015625, + "learning_rate": 5.205074053101306e-05, + "loss": 2.3617, "step": 1784 }, { - "epoch": 0.5716573258606885, - "grad_norm": 0.423828125, - "learning_rate": 2.3114487897577614e-05, - "loss": 2.1639, + "epoch": 0.6935237123916369, + "grad_norm": 0.65234375, + "learning_rate": 5.193169651934603e-05, + "loss": 2.3791, "step": 1785 }, { - "epoch": 0.5719775820656525, - "grad_norm": 0.58203125, - "learning_rate": 2.3086608655148447e-05, - "loss": 2.3979, + "epoch": 0.6939122410820524, + "grad_norm": 0.6328125, + "learning_rate": 5.181274103075215e-05, + "loss": 2.3141, "step": 1786 }, { - "epoch": 0.5722978382706165, - "grad_norm": 0.427734375, - "learning_rate": 2.305873180603519e-05, - "loss": 1.7269, + "epoch": 0.6943007697724679, + "grad_norm": 0.61328125, + "learning_rate": 5.169387428430108e-05, + "loss": 2.3242, "step": 1787 }, { - "epoch": 0.5726180944755804, - "grad_norm": 0.484375, - "learning_rate": 2.303085738510686e-05, - "loss": 2.2162, + "epoch": 0.6946892984628834, + "grad_norm": 0.6796875, + "learning_rate": 5.1575096498899e-05, + "loss": 2.3407, "step": 1788 }, { - "epoch": 0.5729383506805444, - "grad_norm": 0.392578125, - "learning_rate": 2.3002985427229472e-05, - "loss": 2.1375, + "epoch": 0.6950778271532988, + "grad_norm": 0.625, + "learning_rate": 5.1456407893288275e-05, + "loss": 2.3777, "step": 1789 }, { - "epoch": 0.5732586068855084, - "grad_norm": 0.58984375, - "learning_rate": 2.2975115967265924e-05, - "loss": 1.917, + "epoch": 0.6954663558437143, + "grad_norm": 0.64453125, + "learning_rate": 5.133780868604705e-05, + "loss": 2.3728, "step": 1790 }, { - "epoch": 0.5735788630904723, - "grad_norm": 0.4140625, - "learning_rate": 2.294724904007602e-05, - "loss": 2.2711, + "epoch": 0.6958548845341298, + "grad_norm": 0.6484375, + "learning_rate": 5.121929909558887e-05, + "loss": 2.3792, "step": 1791 }, { - "epoch": 0.5738991192954364, - "grad_norm": 0.490234375, - "learning_rate": 2.2919384680516382e-05, - "loss": 2.324, + "epoch": 0.6962434132245453, + "grad_norm": 0.625, + "learning_rate": 5.110087934016209e-05, + "loss": 2.3423, "step": 1792 }, { - "epoch": 0.5742193755004004, - "grad_norm": 0.8671875, - "learning_rate": 2.2891522923440408e-05, - "loss": 2.2216, + "epoch": 0.6966319419149608, + "grad_norm": 0.625, + "learning_rate": 5.098254963784979e-05, + "loss": 2.3595, "step": 1793 }, { - "epoch": 0.5745396317053643, - "grad_norm": 0.58203125, - "learning_rate": 2.286366380369827e-05, - "loss": 1.962, + "epoch": 0.6970204706053763, + "grad_norm": 0.6328125, + "learning_rate": 5.0864310206569144e-05, + "loss": 2.3816, "step": 1794 }, { - "epoch": 0.5748598879103283, - "grad_norm": 0.53125, - "learning_rate": 2.2835807356136817e-05, - "loss": 2.2255, + "epoch": 0.6974089992957917, + "grad_norm": 0.671875, + "learning_rate": 5.074616126407106e-05, + "loss": 2.4156, "step": 1795 }, { - "epoch": 0.5751801441152923, - "grad_norm": 0.53125, - "learning_rate": 2.2807953615599552e-05, - "loss": 1.8936, + "epoch": 0.6977975279862072, + "grad_norm": 0.671875, + "learning_rate": 5.0628103027939855e-05, + "loss": 2.3937, "step": 1796 }, { - "epoch": 0.5755004003202562, - "grad_norm": 0.50390625, - "learning_rate": 2.2780102616926633e-05, - "loss": 1.6788, + "epoch": 0.6981860566766227, + "grad_norm": 0.65625, + "learning_rate": 5.0510135715592745e-05, + "loss": 2.388, "step": 1797 }, { - "epoch": 0.5758206565252202, - "grad_norm": 0.5390625, - "learning_rate": 2.2752254394954736e-05, - "loss": 2.1119, + "epoch": 0.6985745853670382, + "grad_norm": 0.66796875, + "learning_rate": 5.039225954427953e-05, + "loss": 2.3512, "step": 1798 }, { - "epoch": 0.5761409127301842, - "grad_norm": 0.58203125, - "learning_rate": 2.272440898451708e-05, - "loss": 2.2388, + "epoch": 0.6989631140574537, + "grad_norm": 0.640625, + "learning_rate": 5.027447473108218e-05, + "loss": 2.4035, "step": 1799 }, { - "epoch": 0.5764611689351481, - "grad_norm": 0.5078125, - "learning_rate": 2.2696566420443403e-05, - "loss": 2.1596, + "epoch": 0.6993516427478692, + "grad_norm": 0.69921875, + "learning_rate": 5.015678149291439e-05, + "loss": 2.3672, "step": 1800 }, { - "epoch": 0.5767814251401121, - "grad_norm": 0.4765625, - "learning_rate": 2.266872673755983e-05, - "loss": 2.5837, + "epoch": 0.6997401714382846, + "grad_norm": 0.61328125, + "learning_rate": 5.003918004652117e-05, + "loss": 2.3271, "step": 1801 }, { - "epoch": 0.5771016813450761, - "grad_norm": 0.5859375, - "learning_rate": 2.2640889970688904e-05, - "loss": 1.9443, + "epoch": 0.7001287001287001, + "grad_norm": 0.60546875, + "learning_rate": 4.992167060847858e-05, + "loss": 2.3346, "step": 1802 }, { - "epoch": 0.57742193755004, - "grad_norm": 0.5546875, - "learning_rate": 2.2613056154649542e-05, - "loss": 2.2111, + "epoch": 0.7005172288191156, + "grad_norm": 0.625, + "learning_rate": 4.9804253395193154e-05, + "loss": 2.2578, "step": 1803 }, { - "epoch": 0.577742193755004, - "grad_norm": 0.5703125, - "learning_rate": 2.2585225324256928e-05, - "loss": 2.4131, + "epoch": 0.7009057575095311, + "grad_norm": 0.640625, + "learning_rate": 4.968692862290165e-05, + "loss": 2.3872, "step": 1804 }, { - "epoch": 0.578062449959968, - "grad_norm": 0.5546875, - "learning_rate": 2.255739751432256e-05, - "loss": 2.4771, + "epoch": 0.7012942861999466, + "grad_norm": 0.65625, + "learning_rate": 4.9569696507670535e-05, + "loss": 2.3211, "step": 1805 }, { - "epoch": 0.5783827061649319, - "grad_norm": 0.458984375, - "learning_rate": 2.252957275965411e-05, - "loss": 2.0832, + "epoch": 0.7016828148903621, + "grad_norm": 0.63671875, + "learning_rate": 4.945255726539566e-05, + "loss": 2.3605, "step": 1806 }, { - "epoch": 0.5787029623698959, - "grad_norm": 0.6171875, - "learning_rate": 2.2501751095055448e-05, - "loss": 2.2442, + "epoch": 0.7020713435807775, + "grad_norm": 0.66015625, + "learning_rate": 4.9335511111801836e-05, + "loss": 2.4147, "step": 1807 }, { - "epoch": 0.5790232185748598, - "grad_norm": 0.41015625, - "learning_rate": 2.2473932555326604e-05, - "loss": 2.3033, + "epoch": 0.702459872271193, + "grad_norm": 0.66015625, + "learning_rate": 4.921855826244248e-05, + "loss": 2.3586, "step": 1808 }, { - "epoch": 0.5793434747798238, - "grad_norm": 0.423828125, - "learning_rate": 2.2446117175263658e-05, - "loss": 2.1901, + "epoch": 0.7028484009616085, + "grad_norm": 0.60546875, + "learning_rate": 4.910169893269908e-05, + "loss": 2.279, "step": 1809 }, { - "epoch": 0.5796637309847879, - "grad_norm": 3.03125, - "learning_rate": 2.2418304989658757e-05, - "loss": 2.1636, + "epoch": 0.703236929652024, + "grad_norm": 0.609375, + "learning_rate": 4.8984933337781e-05, + "loss": 2.3442, "step": 1810 }, { - "epoch": 0.5799839871897517, - "grad_norm": 0.416015625, - "learning_rate": 2.239049603330006e-05, - "loss": 2.5267, + "epoch": 0.7036254583424395, + "grad_norm": 0.7109375, + "learning_rate": 4.886826169272489e-05, + "loss": 2.3817, "step": 1811 }, { - "epoch": 0.5803042433947158, - "grad_norm": 0.443359375, - "learning_rate": 2.2362690340971678e-05, - "loss": 1.9348, + "epoch": 0.704013987032855, + "grad_norm": 0.62109375, + "learning_rate": 4.8751684212394466e-05, + "loss": 2.3372, "step": 1812 }, { - "epoch": 0.5806244995996798, - "grad_norm": 1.0234375, - "learning_rate": 2.233488794745362e-05, - "loss": 2.3273, + "epoch": 0.7044025157232704, + "grad_norm": 0.640625, + "learning_rate": 4.8635201111479986e-05, + "loss": 2.403, "step": 1813 }, { - "epoch": 0.5809447558046437, - "grad_norm": 0.69921875, - "learning_rate": 2.2307088887521816e-05, - "loss": 1.9374, + "epoch": 0.7047910444136859, + "grad_norm": 0.62890625, + "learning_rate": 4.851881260449791e-05, + "loss": 2.3849, "step": 1814 }, { - "epoch": 0.5812650120096077, - "grad_norm": 0.5234375, - "learning_rate": 2.2279293195947977e-05, - "loss": 2.5428, + "epoch": 0.7051795731041014, + "grad_norm": 0.640625, + "learning_rate": 4.840251890579047e-05, + "loss": 2.3893, "step": 1815 }, { - "epoch": 0.5815852682145717, - "grad_norm": 0.43359375, - "learning_rate": 2.225150090749964e-05, - "loss": 1.9329, + "epoch": 0.7055681017945169, + "grad_norm": 0.671875, + "learning_rate": 4.828632022952535e-05, + "loss": 2.3303, "step": 1816 }, { - "epoch": 0.5819055244195356, - "grad_norm": 0.484375, - "learning_rate": 2.222371205694006e-05, - "loss": 2.0871, + "epoch": 0.7059566304849324, + "grad_norm": 0.6328125, + "learning_rate": 4.817021678969518e-05, + "loss": 2.3204, "step": 1817 }, { - "epoch": 0.5822257806244996, - "grad_norm": 0.78125, - "learning_rate": 2.219592667902819e-05, - "loss": 2.1848, + "epoch": 0.7063451591753479, + "grad_norm": 0.61328125, + "learning_rate": 4.805420880011723e-05, + "loss": 2.3551, "step": 1818 }, { - "epoch": 0.5825460368294636, - "grad_norm": 0.51171875, - "learning_rate": 2.2168144808518667e-05, - "loss": 1.9535, + "epoch": 0.7067336878657633, + "grad_norm": 0.6484375, + "learning_rate": 4.793829647443302e-05, + "loss": 2.3654, "step": 1819 }, { - "epoch": 0.5828662930344275, - "grad_norm": 0.447265625, - "learning_rate": 2.2140366480161713e-05, - "loss": 2.2323, + "epoch": 0.7071222165561788, + "grad_norm": 0.6328125, + "learning_rate": 4.782248002610781e-05, + "loss": 2.2714, "step": 1820 }, { - "epoch": 0.5831865492393915, - "grad_norm": 0.63671875, - "learning_rate": 2.211259172870312e-05, - "loss": 2.3016, + "epoch": 0.7075107452465943, + "grad_norm": 0.640625, + "learning_rate": 4.770675966843035e-05, + "loss": 2.3196, "step": 1821 }, { - "epoch": 0.5835068054443555, - "grad_norm": 0.392578125, - "learning_rate": 2.2084820588884234e-05, - "loss": 1.6852, + "epoch": 0.7078992739370098, + "grad_norm": 0.625, + "learning_rate": 4.7591135614512417e-05, + "loss": 2.4134, "step": 1822 }, { - "epoch": 0.5838270616493194, - "grad_norm": 2.328125, - "learning_rate": 2.2057053095441854e-05, - "loss": 2.3596, + "epoch": 0.7082878026274253, + "grad_norm": 0.62890625, + "learning_rate": 4.747560807728847e-05, + "loss": 2.3107, "step": 1823 }, { - "epoch": 0.5841473178542834, - "grad_norm": 0.392578125, - "learning_rate": 2.202928928310822e-05, - "loss": 1.8076, + "epoch": 0.7086763313178408, + "grad_norm": 0.61328125, + "learning_rate": 4.736017726951515e-05, + "loss": 2.3464, "step": 1824 }, { - "epoch": 0.5844675740592474, - "grad_norm": 0.50390625, - "learning_rate": 2.2001529186610996e-05, - "loss": 2.6958, + "epoch": 0.7090648600082562, + "grad_norm": 0.64453125, + "learning_rate": 4.724484340377102e-05, + "loss": 2.3962, "step": 1825 }, { - "epoch": 0.5847878302642113, - "grad_norm": 0.77734375, - "learning_rate": 2.197377284067317e-05, - "loss": 2.3731, + "epoch": 0.7094533886986717, + "grad_norm": 0.6640625, + "learning_rate": 4.7129606692456095e-05, + "loss": 2.411, "step": 1826 }, { - "epoch": 0.5851080864691753, - "grad_norm": 0.53125, - "learning_rate": 2.194602028001303e-05, - "loss": 2.4798, + "epoch": 0.7098419173890872, + "grad_norm": 0.64453125, + "learning_rate": 4.701446734779149e-05, + "loss": 2.3093, "step": 1827 }, { - "epoch": 0.5854283426741393, - "grad_norm": 0.5625, - "learning_rate": 2.1918271539344177e-05, - "loss": 2.7211, + "epoch": 0.7102304460795027, + "grad_norm": 0.6328125, + "learning_rate": 4.6899425581818924e-05, + "loss": 2.3815, "step": 1828 }, { - "epoch": 0.5857485988791032, - "grad_norm": 0.482421875, - "learning_rate": 2.1890526653375372e-05, - "loss": 2.6734, + "epoch": 0.7106189747699182, + "grad_norm": 0.6875, + "learning_rate": 4.6784481606400496e-05, + "loss": 2.4293, "step": 1829 }, { - "epoch": 0.5860688550840673, - "grad_norm": 0.62890625, - "learning_rate": 2.186278565681061e-05, - "loss": 2.3266, + "epoch": 0.7110075034603337, + "grad_norm": 0.65234375, + "learning_rate": 4.66696356332182e-05, + "loss": 2.3517, "step": 1830 }, { - "epoch": 0.5863891112890313, - "grad_norm": 1.1015625, - "learning_rate": 2.183504858434899e-05, - "loss": 2.572, + "epoch": 0.7113960321507491, + "grad_norm": 0.67578125, + "learning_rate": 4.655488787377354e-05, + "loss": 2.4058, "step": 1831 }, { - "epoch": 0.5867093674939952, - "grad_norm": 0.59765625, - "learning_rate": 2.1807315470684697e-05, - "loss": 2.4207, + "epoch": 0.7117845608411646, + "grad_norm": 0.6171875, + "learning_rate": 4.644023853938716e-05, + "loss": 2.2576, "step": 1832 }, { - "epoch": 0.5870296236989592, - "grad_norm": 0.74609375, - "learning_rate": 2.1779586350506996e-05, - "loss": 2.1545, + "epoch": 0.7121730895315801, + "grad_norm": 0.6015625, + "learning_rate": 4.6325687841198415e-05, + "loss": 2.2638, "step": 1833 }, { - "epoch": 0.5873498799039232, - "grad_norm": 0.447265625, - "learning_rate": 2.175186125850012e-05, - "loss": 2.287, + "epoch": 0.7125616182219956, + "grad_norm": 0.65234375, + "learning_rate": 4.621123599016505e-05, + "loss": 2.2876, "step": 1834 }, { - "epoch": 0.5876701361088871, - "grad_norm": 0.333984375, - "learning_rate": 2.1724140229343288e-05, - "loss": 2.3484, + "epoch": 0.7129501469124111, + "grad_norm": 0.640625, + "learning_rate": 4.609688319706271e-05, + "loss": 2.316, "step": 1835 }, { - "epoch": 0.5879903923138511, - "grad_norm": 0.478515625, - "learning_rate": 2.169642329771063e-05, - "loss": 2.0906, + "epoch": 0.7133386756028266, + "grad_norm": 0.66015625, + "learning_rate": 4.5982629672484734e-05, + "loss": 2.3727, "step": 1836 }, { - "epoch": 0.5883106485188151, - "grad_norm": 0.51953125, - "learning_rate": 2.1668710498271162e-05, - "loss": 2.4591, + "epoch": 0.713727204293242, + "grad_norm": 0.66015625, + "learning_rate": 4.586847562684147e-05, + "loss": 2.3608, "step": 1837 }, { - "epoch": 0.588630904723779, - "grad_norm": 0.52734375, - "learning_rate": 2.1641001865688686e-05, - "loss": 1.9441, + "epoch": 0.7141157329836575, + "grad_norm": 0.63671875, + "learning_rate": 4.575442127036019e-05, + "loss": 2.2822, "step": 1838 }, { - "epoch": 0.588951160928743, - "grad_norm": 0.51171875, - "learning_rate": 2.1613297434621865e-05, - "loss": 2.4205, + "epoch": 0.714504261674073, + "grad_norm": 0.6328125, + "learning_rate": 4.564046681308457e-05, + "loss": 2.3288, "step": 1839 }, { - "epoch": 0.589271417133707, - "grad_norm": 0.6484375, - "learning_rate": 2.158559723972404e-05, - "loss": 2.1306, + "epoch": 0.7148927903644885, + "grad_norm": 0.73046875, + "learning_rate": 4.5526612464874285e-05, + "loss": 2.3741, "step": 1840 }, { - "epoch": 0.5895916733386709, - "grad_norm": 0.400390625, - "learning_rate": 2.15579013156433e-05, - "loss": 1.831, + "epoch": 0.715281319054904, + "grad_norm": 0.6328125, + "learning_rate": 4.541285843540465e-05, + "loss": 2.2713, "step": 1841 }, { - "epoch": 0.5899119295436349, - "grad_norm": 0.5546875, - "learning_rate": 2.1530209697022366e-05, - "loss": 1.9654, + "epoch": 0.7156698477453195, + "grad_norm": 0.61328125, + "learning_rate": 4.5299204934166216e-05, + "loss": 2.3915, "step": 1842 }, { - "epoch": 0.5902321857485989, - "grad_norm": 0.5078125, - "learning_rate": 2.150252241849857e-05, - "loss": 2.4949, + "epoch": 0.7160583764357349, + "grad_norm": 0.64453125, + "learning_rate": 4.518565217046445e-05, + "loss": 2.3431, "step": 1843 }, { - "epoch": 0.5905524419535628, - "grad_norm": 0.64453125, - "learning_rate": 2.1474839514703843e-05, - "loss": 1.785, + "epoch": 0.7164469051261504, + "grad_norm": 0.6328125, + "learning_rate": 4.507220035341929e-05, + "loss": 2.3615, "step": 1844 }, { - "epoch": 0.5908726981585268, - "grad_norm": 0.5078125, - "learning_rate": 2.1447161020264616e-05, - "loss": 1.4549, + "epoch": 0.7168354338165659, + "grad_norm": 0.6171875, + "learning_rate": 4.4958849691964675e-05, + "loss": 2.2747, "step": 1845 }, { - "epoch": 0.5911929543634908, - "grad_norm": 0.54296875, - "learning_rate": 2.141948696980181e-05, - "loss": 2.0587, + "epoch": 0.7172239625069814, + "grad_norm": 0.66015625, + "learning_rate": 4.484560039484836e-05, + "loss": 2.359, "step": 1846 }, { - "epoch": 0.5915132105684547, - "grad_norm": 0.54296875, - "learning_rate": 2.1391817397930814e-05, - "loss": 1.8139, + "epoch": 0.7176124911973969, + "grad_norm": 0.6875, + "learning_rate": 4.47324526706314e-05, + "loss": 2.4031, "step": 1847 }, { - "epoch": 0.5918334667734187, - "grad_norm": 0.458984375, - "learning_rate": 2.1364152339261374e-05, - "loss": 2.3478, + "epoch": 0.7180010198878123, + "grad_norm": 0.66015625, + "learning_rate": 4.4619406727687785e-05, + "loss": 2.4118, "step": 1848 }, { - "epoch": 0.5921537229783828, - "grad_norm": 0.90625, - "learning_rate": 2.133649182839761e-05, - "loss": 1.9641, + "epoch": 0.7183895485782278, + "grad_norm": 3.046875, + "learning_rate": 4.450646277420407e-05, + "loss": 2.4098, "step": 1849 }, { - "epoch": 0.5924739791833467, - "grad_norm": 0.52734375, - "learning_rate": 2.1308835899937972e-05, - "loss": 2.3559, + "epoch": 0.7187780772686433, + "grad_norm": 0.61328125, + "learning_rate": 4.439362101817898e-05, + "loss": 2.3236, "step": 1850 }, { - "epoch": 0.5927942353883107, - "grad_norm": 0.546875, - "learning_rate": 2.128118458847515e-05, - "loss": 2.3271, + "epoch": 0.7191666059590588, + "grad_norm": 0.62109375, + "learning_rate": 4.4280881667423026e-05, + "loss": 2.3833, "step": 1851 }, { - "epoch": 0.5931144915932747, - "grad_norm": 0.478515625, - "learning_rate": 2.1253537928596057e-05, - "loss": 2.0447, + "epoch": 0.7195551346494743, + "grad_norm": 0.65234375, + "learning_rate": 4.416824492955817e-05, + "loss": 2.3257, "step": 1852 }, { - "epoch": 0.5934347477982386, - "grad_norm": 0.50390625, - "learning_rate": 2.1225895954881823e-05, - "loss": 2.2835, + "epoch": 0.7199436633398898, + "grad_norm": 0.65625, + "learning_rate": 4.40557110120173e-05, + "loss": 2.3589, "step": 1853 }, { - "epoch": 0.5937550040032026, - "grad_norm": 0.53515625, - "learning_rate": 2.119825870190767e-05, - "loss": 1.9083, + "epoch": 0.7203321920303052, + "grad_norm": 0.65625, + "learning_rate": 4.394328012204409e-05, + "loss": 2.357, "step": 1854 }, { - "epoch": 0.5940752602081665, - "grad_norm": 0.5234375, - "learning_rate": 2.1170626204242962e-05, - "loss": 2.0059, + "epoch": 0.7207207207207207, + "grad_norm": 0.6796875, + "learning_rate": 4.383095246669232e-05, + "loss": 2.3558, "step": 1855 }, { - "epoch": 0.5943955164131305, - "grad_norm": 0.6015625, - "learning_rate": 2.114299849645108e-05, - "loss": 2.408, + "epoch": 0.7211092494111362, + "grad_norm": 0.6328125, + "learning_rate": 4.371872825282578e-05, + "loss": 2.3796, "step": 1856 }, { - "epoch": 0.5947157726180945, - "grad_norm": 0.55859375, - "learning_rate": 2.1115375613089416e-05, - "loss": 2.1296, + "epoch": 0.7214977781015517, + "grad_norm": 0.6796875, + "learning_rate": 4.3606607687117704e-05, + "loss": 2.4108, "step": 1857 }, { - "epoch": 0.5950360288230584, - "grad_norm": 0.5703125, - "learning_rate": 2.1087757588709356e-05, - "loss": 2.2477, + "epoch": 0.7218863067919672, + "grad_norm": 0.640625, + "learning_rate": 4.349459097605047e-05, + "loss": 2.3564, "step": 1858 }, { - "epoch": 0.5953562850280224, - "grad_norm": 0.578125, - "learning_rate": 2.106014445785618e-05, - "loss": 2.5451, + "epoch": 0.7222748354823827, + "grad_norm": 0.62109375, + "learning_rate": 4.338267832591519e-05, + "loss": 2.3506, "step": 1859 }, { - "epoch": 0.5956765412329864, - "grad_norm": 0.5, - "learning_rate": 2.1032536255069045e-05, - "loss": 1.9078, + "epoch": 0.7226633641727981, + "grad_norm": 0.6640625, + "learning_rate": 4.327086994281132e-05, + "loss": 2.3364, "step": 1860 }, { - "epoch": 0.5959967974379503, - "grad_norm": 0.5859375, - "learning_rate": 2.1004933014880976e-05, - "loss": 2.3821, + "epoch": 0.7230518928632136, + "grad_norm": 0.625, + "learning_rate": 4.31591660326463e-05, + "loss": 2.4074, "step": 1861 }, { - "epoch": 0.5963170536429143, - "grad_norm": 0.310546875, - "learning_rate": 2.097733477181876e-05, - "loss": 1.7438, + "epoch": 0.7234404215536291, + "grad_norm": 3.0, + "learning_rate": 4.30475668011352e-05, + "loss": 2.3424, "step": 1862 }, { - "epoch": 0.5966373098478783, - "grad_norm": 0.61328125, - "learning_rate": 2.094974156040293e-05, - "loss": 1.9844, + "epoch": 0.7238289502440446, + "grad_norm": 0.640625, + "learning_rate": 4.29360724538003e-05, + "loss": 2.3817, "step": 1863 }, { - "epoch": 0.5969575660528422, - "grad_norm": 0.57421875, - "learning_rate": 2.092215341514776e-05, - "loss": 2.3406, + "epoch": 0.7242174789344601, + "grad_norm": 0.64453125, + "learning_rate": 4.282468319597067e-05, + "loss": 2.3692, "step": 1864 }, { - "epoch": 0.5972778222578062, - "grad_norm": 0.474609375, - "learning_rate": 2.0894570370561156e-05, - "loss": 2.2888, + "epoch": 0.7246060076248756, + "grad_norm": 0.65625, + "learning_rate": 4.2713399232781914e-05, + "loss": 2.2731, "step": 1865 }, { - "epoch": 0.5975980784627702, - "grad_norm": 0.5234375, - "learning_rate": 2.0866992461144645e-05, - "loss": 2.4423, + "epoch": 0.724994536315291, + "grad_norm": 0.63671875, + "learning_rate": 4.2602220769175717e-05, + "loss": 2.3326, "step": 1866 }, { - "epoch": 0.5979183346677341, - "grad_norm": 0.490234375, - "learning_rate": 2.0839419721393354e-05, - "loss": 2.4976, + "epoch": 0.7253830650057065, + "grad_norm": 0.65234375, + "learning_rate": 4.249114800989944e-05, + "loss": 2.389, "step": 1867 }, { - "epoch": 0.5982385908726982, - "grad_norm": 0.431640625, - "learning_rate": 2.0811852185795916e-05, - "loss": 2.1009, + "epoch": 0.725771593696122, + "grad_norm": 0.625, + "learning_rate": 4.238018115950582e-05, + "loss": 2.2927, "step": 1868 }, { - "epoch": 0.5985588470776622, - "grad_norm": 0.41015625, - "learning_rate": 2.078428988883448e-05, - "loss": 2.1664, + "epoch": 0.7261601223865375, + "grad_norm": 0.62890625, + "learning_rate": 4.22693204223525e-05, + "loss": 2.3935, "step": 1869 }, { - "epoch": 0.5988791032826261, + "epoch": 0.726548651076953, "grad_norm": 0.59765625, - "learning_rate": 2.075673286498463e-05, - "loss": 2.3979, + "learning_rate": 4.215856600260176e-05, + "loss": 2.3794, "step": 1870 }, { - "epoch": 0.5991993594875901, - "grad_norm": 0.451171875, - "learning_rate": 2.0729181148715332e-05, - "loss": 1.8152, + "epoch": 0.7269371797673685, + "grad_norm": 0.6640625, + "learning_rate": 4.204791810422008e-05, + "loss": 2.3499, "step": 1871 }, { - "epoch": 0.5995196156925541, - "grad_norm": 0.494140625, - "learning_rate": 2.070163477448896e-05, - "loss": 2.2508, + "epoch": 0.7273257084577839, + "grad_norm": 0.62109375, + "learning_rate": 4.1937376930977704e-05, + "loss": 2.3803, "step": 1872 }, { - "epoch": 0.599839871897518, - "grad_norm": 0.4921875, - "learning_rate": 2.0674093776761176e-05, - "loss": 2.1103, + "epoch": 0.7277142371481994, + "grad_norm": 0.63671875, + "learning_rate": 4.182694268644837e-05, + "loss": 2.3551, "step": 1873 }, { - "epoch": 0.600160128102482, - "grad_norm": 0.5078125, - "learning_rate": 2.064655818998091e-05, - "loss": 2.4928, + "epoch": 0.7281027658386149, + "grad_norm": 0.58984375, + "learning_rate": 4.17166155740089e-05, + "loss": 2.3986, "step": 1874 }, { - "epoch": 0.600480384307446, - "grad_norm": 0.58203125, - "learning_rate": 2.061902804859035e-05, - "loss": 2.1205, + "epoch": 0.7284912945290304, + "grad_norm": 0.60546875, + "learning_rate": 4.160639579683885e-05, + "loss": 2.301, "step": 1875 }, { - "epoch": 0.6008006405124099, - "grad_norm": 0.59375, - "learning_rate": 2.0591503387024853e-05, - "loss": 2.2769, + "epoch": 0.7288798232194459, + "grad_norm": 0.62109375, + "learning_rate": 4.149628355792005e-05, + "loss": 2.3123, "step": 1876 }, { - "epoch": 0.6011208967173739, - "grad_norm": 0.57421875, - "learning_rate": 2.056398423971292e-05, - "loss": 2.2979, + "epoch": 0.7292683519098614, + "grad_norm": 0.62109375, + "learning_rate": 4.1386279060036316e-05, + "loss": 2.3621, "step": 1877 }, { - "epoch": 0.6014411529223379, - "grad_norm": 0.318359375, - "learning_rate": 2.053647064107618e-05, - "loss": 1.9518, + "epoch": 0.7296568806002768, + "grad_norm": 0.62109375, + "learning_rate": 4.127638250577305e-05, + "loss": 2.305, "step": 1878 }, { - "epoch": 0.6017614091273018, - "grad_norm": 0.46875, - "learning_rate": 2.0508962625529284e-05, - "loss": 2.1684, + "epoch": 0.7300454092906923, + "grad_norm": 0.625, + "learning_rate": 4.1166594097516854e-05, + "loss": 2.3822, "step": 1879 }, { - "epoch": 0.6020816653322658, - "grad_norm": 0.51171875, - "learning_rate": 2.0481460227479942e-05, - "loss": 2.4742, + "epoch": 0.7304339379811078, + "grad_norm": 0.62890625, + "learning_rate": 4.105691403745519e-05, + "loss": 2.346, "step": 1880 }, { - "epoch": 0.6024019215372298, - "grad_norm": 0.439453125, - "learning_rate": 2.0453963481328796e-05, - "loss": 2.1223, + "epoch": 0.7308224666715233, + "grad_norm": 0.64453125, + "learning_rate": 4.0947342527575916e-05, + "loss": 2.3148, "step": 1881 }, { - "epoch": 0.6027221777421937, - "grad_norm": 0.55078125, - "learning_rate": 2.042647242146944e-05, - "loss": 1.8289, + "epoch": 0.7312109953619388, + "grad_norm": 0.61328125, + "learning_rate": 4.0837879769667045e-05, + "loss": 2.3351, "step": 1882 }, { - "epoch": 0.6030424339471577, - "grad_norm": 0.3828125, - "learning_rate": 2.0398987082288363e-05, - "loss": 2.206, + "epoch": 0.7315995240523543, + "grad_norm": 0.5859375, + "learning_rate": 4.072852596531631e-05, + "loss": 2.3522, "step": 1883 }, { - "epoch": 0.6033626901521217, - "grad_norm": 0.431640625, - "learning_rate": 2.037150749816488e-05, - "loss": 1.9582, + "epoch": 0.7319880527427697, + "grad_norm": 0.67578125, + "learning_rate": 4.06192813159108e-05, + "loss": 2.4085, "step": 1884 }, { - "epoch": 0.6036829463570856, - "grad_norm": 0.56640625, - "learning_rate": 2.0344033703471106e-05, - "loss": 2.1088, + "epoch": 0.7323765814331852, + "grad_norm": 0.62890625, + "learning_rate": 4.051014602263652e-05, + "loss": 2.3633, "step": 1885 }, { - "epoch": 0.6040032025620496, - "grad_norm": 0.359375, - "learning_rate": 2.0316565732571947e-05, - "loss": 1.6995, + "epoch": 0.7327651101236007, + "grad_norm": 0.64453125, + "learning_rate": 4.0401120286478154e-05, + "loss": 2.422, "step": 1886 }, { - "epoch": 0.6043234587670137, - "grad_norm": 0.63671875, - "learning_rate": 2.028910361982499e-05, - "loss": 2.3343, + "epoch": 0.7331536388140162, + "grad_norm": 0.625, + "learning_rate": 4.029220430821862e-05, + "loss": 2.4292, "step": 1887 }, { - "epoch": 0.6046437149719776, - "grad_norm": 0.44140625, - "learning_rate": 2.0261647399580493e-05, - "loss": 2.5429, + "epoch": 0.7335421675044317, + "grad_norm": 0.6015625, + "learning_rate": 4.018339828843868e-05, + "loss": 2.3436, "step": 1888 }, { - "epoch": 0.6049639711769416, - "grad_norm": 0.48046875, - "learning_rate": 2.023419710618138e-05, - "loss": 2.6612, + "epoch": 0.7339306961948472, + "grad_norm": 0.625, + "learning_rate": 4.007470242751654e-05, + "loss": 2.4115, "step": 1889 }, { - "epoch": 0.6052842273819056, - "grad_norm": 1.140625, - "learning_rate": 2.0206752773963128e-05, - "loss": 2.0385, + "epoch": 0.7343192248852626, + "grad_norm": 0.625, + "learning_rate": 3.996611692562764e-05, + "loss": 2.4112, "step": 1890 }, { - "epoch": 0.6056044835868695, - "grad_norm": 0.462890625, - "learning_rate": 2.0179314437253756e-05, - "loss": 1.9074, + "epoch": 0.7347077535756781, + "grad_norm": 0.625, + "learning_rate": 3.985764198274414e-05, + "loss": 2.3491, "step": 1891 }, { - "epoch": 0.6059247397918335, - "grad_norm": 0.54296875, - "learning_rate": 2.015188213037381e-05, - "loss": 2.1805, + "epoch": 0.7350962822660936, + "grad_norm": 0.60546875, + "learning_rate": 3.9749277798634575e-05, + "loss": 2.288, "step": 1892 }, { - "epoch": 0.6062449959967975, - "grad_norm": 0.4765625, - "learning_rate": 2.012445588763627e-05, - "loss": 2.714, + "epoch": 0.7354848109565091, + "grad_norm": 0.6171875, + "learning_rate": 3.964102457286353e-05, + "loss": 2.3784, "step": 1893 }, { - "epoch": 0.6065652522017614, - "grad_norm": 0.498046875, - "learning_rate": 2.009703574334656e-05, - "loss": 2.5523, + "epoch": 0.7358733396469246, + "grad_norm": 0.6640625, + "learning_rate": 3.9532882504791236e-05, + "loss": 2.3957, "step": 1894 }, { - "epoch": 0.6068855084067254, - "grad_norm": 0.451171875, - "learning_rate": 2.0069621731802435e-05, - "loss": 2.4568, + "epoch": 0.7362618683373401, + "grad_norm": 0.625, + "learning_rate": 3.9424851793573224e-05, + "loss": 2.358, "step": 1895 }, { - "epoch": 0.6072057646116894, - "grad_norm": 0.388671875, - "learning_rate": 2.0042213887293994e-05, - "loss": 2.021, + "epoch": 0.7366503970277555, + "grad_norm": 0.62109375, + "learning_rate": 3.931693263815997e-05, + "loss": 2.4146, "step": 1896 }, { - "epoch": 0.6075260208166533, - "grad_norm": 1.1484375, - "learning_rate": 2.0014812244103653e-05, - "loss": 2.636, + "epoch": 0.737038925718171, + "grad_norm": 0.65625, + "learning_rate": 3.920912523729642e-05, + "loss": 2.4071, "step": 1897 }, { - "epoch": 0.6078462770216173, - "grad_norm": 0.498046875, - "learning_rate": 1.9987416836506034e-05, - "loss": 1.8559, + "epoch": 0.7374274544085865, + "grad_norm": 0.69921875, + "learning_rate": 3.9101429789521827e-05, + "loss": 2.2899, "step": 1898 }, { - "epoch": 0.6081665332265813, - "grad_norm": 0.53515625, - "learning_rate": 1.9960027698767957e-05, - "loss": 1.9964, + "epoch": 0.737815983099002, + "grad_norm": 0.66015625, + "learning_rate": 3.8993846493169164e-05, + "loss": 2.3797, "step": 1899 }, { - "epoch": 0.6084867894315452, - "grad_norm": 0.4453125, - "learning_rate": 1.993264486514844e-05, - "loss": 2.045, + "epoch": 0.7382045117894175, + "grad_norm": 0.625, + "learning_rate": 3.888637554636493e-05, + "loss": 2.4159, "step": 1900 }, { - "epoch": 0.6088070456365092, - "grad_norm": 1.140625, - "learning_rate": 1.9905268369898576e-05, - "loss": 2.0631, + "epoch": 0.738593040479833, + "grad_norm": 0.67578125, + "learning_rate": 3.877901714702873e-05, + "loss": 2.385, "step": 1901 }, { - "epoch": 0.6091273018414731, - "grad_norm": 0.470703125, - "learning_rate": 1.9877898247261534e-05, - "loss": 1.8805, + "epoch": 0.7389815691702484, + "grad_norm": 0.6484375, + "learning_rate": 3.867177149287286e-05, + "loss": 2.2841, "step": 1902 }, { - "epoch": 0.6094475580464371, - "grad_norm": 0.5234375, - "learning_rate": 1.9850534531472546e-05, - "loss": 2.0472, + "epoch": 0.7393700978606639, + "grad_norm": 0.609375, + "learning_rate": 3.8564638781402e-05, + "loss": 2.3939, "step": 1903 }, { - "epoch": 0.6097678142514011, - "grad_norm": 0.474609375, - "learning_rate": 1.9823177256758787e-05, - "loss": 2.5672, + "epoch": 0.7397586265510794, + "grad_norm": 0.6171875, + "learning_rate": 3.845761920991283e-05, + "loss": 2.4283, "step": 1904 }, { - "epoch": 0.610088070456365, - "grad_norm": 0.431640625, - "learning_rate": 1.9795826457339393e-05, - "loss": 1.8286, + "epoch": 0.7401471552414949, + "grad_norm": 0.625, + "learning_rate": 3.83507129754937e-05, + "loss": 2.4112, "step": 1905 }, { - "epoch": 0.610408326661329, - "grad_norm": 0.55859375, - "learning_rate": 1.9768482167425413e-05, - "loss": 1.6787, + "epoch": 0.7405356839319104, + "grad_norm": 0.67578125, + "learning_rate": 3.824392027502417e-05, + "loss": 2.3811, "step": 1906 }, { - "epoch": 0.6107285828662931, - "grad_norm": 0.5625, - "learning_rate": 1.974114442121973e-05, - "loss": 2.3295, + "epoch": 0.7409242126223259, + "grad_norm": 0.6328125, + "learning_rate": 3.8137241305174834e-05, + "loss": 2.3265, "step": 1907 }, { - "epoch": 0.611048839071257, - "grad_norm": 0.609375, - "learning_rate": 1.9713813252917075e-05, - "loss": 2.6286, + "epoch": 0.7413127413127413, + "grad_norm": 0.703125, + "learning_rate": 3.803067626240665e-05, + "loss": 2.2776, "step": 1908 }, { - "epoch": 0.611369095276221, + "epoch": 0.7417012700031568, "grad_norm": 0.6796875, - "learning_rate": 1.9686488696703914e-05, - "loss": 2.2819, + "learning_rate": 3.7924225342970945e-05, + "loss": 2.4382, "step": 1909 }, { - "epoch": 0.611689351481185, - "grad_norm": 0.375, - "learning_rate": 1.965917078675845e-05, - "loss": 2.2437, + "epoch": 0.7420897986935723, + "grad_norm": 0.62890625, + "learning_rate": 3.781788874290879e-05, + "loss": 2.3497, "step": 1910 }, { - "epoch": 0.6120096076861489, - "grad_norm": 0.50390625, - "learning_rate": 1.9631859557250605e-05, - "loss": 2.1633, + "epoch": 0.7424783273839878, + "grad_norm": 0.59765625, + "learning_rate": 3.771166665805075e-05, + "loss": 2.3409, "step": 1911 }, { - "epoch": 0.6123298638911129, - "grad_norm": 0.51953125, - "learning_rate": 1.960455504234191e-05, - "loss": 2.1793, + "epoch": 0.7428668560744033, + "grad_norm": 0.60546875, + "learning_rate": 3.760555928401647e-05, + "loss": 2.3056, "step": 1912 }, { - "epoch": 0.6126501200960769, - "grad_norm": 0.515625, - "learning_rate": 1.95772572761855e-05, - "loss": 2.4874, + "epoch": 0.7432553847648187, + "grad_norm": 0.640625, + "learning_rate": 3.749956681621438e-05, + "loss": 2.3703, "step": 1913 }, { - "epoch": 0.6129703763010408, - "grad_norm": 0.53515625, - "learning_rate": 1.9549966292926095e-05, - "loss": 2.1151, + "epoch": 0.7436439134552342, + "grad_norm": 0.62890625, + "learning_rate": 3.7393689449841286e-05, + "loss": 2.3297, "step": 1914 }, { - "epoch": 0.6132906325060048, - "grad_norm": 0.482421875, - "learning_rate": 1.9522682126699903e-05, - "loss": 2.0204, + "epoch": 0.7440324421456497, + "grad_norm": 0.63671875, + "learning_rate": 3.728792737988204e-05, + "loss": 2.3389, "step": 1915 }, { - "epoch": 0.6136108887109688, - "grad_norm": 0.8671875, - "learning_rate": 1.949540481163461e-05, - "loss": 2.5082, + "epoch": 0.7444209708360652, + "grad_norm": 0.6328125, + "learning_rate": 3.718228080110906e-05, + "loss": 2.3029, "step": 1916 }, { - "epoch": 0.6139311449159327, - "grad_norm": 0.41796875, - "learning_rate": 1.9468134381849355e-05, - "loss": 2.2812, + "epoch": 0.7448094995264807, + "grad_norm": 0.64453125, + "learning_rate": 3.707674990808219e-05, + "loss": 2.3986, "step": 1917 }, { - "epoch": 0.6142514011208967, - "grad_norm": 0.515625, - "learning_rate": 1.944087087145462e-05, - "loss": 2.2605, + "epoch": 0.7451980282168962, + "grad_norm": 0.59765625, + "learning_rate": 3.69713348951482e-05, + "loss": 2.2478, "step": 1918 }, { - "epoch": 0.6145716573258607, - "grad_norm": 1.9921875, - "learning_rate": 1.9413614314552293e-05, - "loss": 1.7432, + "epoch": 0.7455865569073116, + "grad_norm": 0.625, + "learning_rate": 3.686603595644044e-05, + "loss": 2.4261, "step": 1919 }, { - "epoch": 0.6148919135308246, - "grad_norm": 0.74609375, - "learning_rate": 1.938636474523551e-05, - "loss": 2.4423, + "epoch": 0.7459750855977271, + "grad_norm": 0.6328125, + "learning_rate": 3.6760853285878486e-05, + "loss": 2.3709, "step": 1920 }, { - "epoch": 0.6152121697357886, - "grad_norm": 0.69140625, - "learning_rate": 1.935912219758868e-05, - "loss": 2.3166, + "epoch": 0.7463636142881426, + "grad_norm": 0.60546875, + "learning_rate": 3.6655787077167825e-05, + "loss": 2.3114, "step": 1921 }, { - "epoch": 0.6155324259407526, - "grad_norm": 0.255859375, - "learning_rate": 1.9331886705687464e-05, - "loss": 1.7659, + "epoch": 0.7467521429785581, + "grad_norm": 0.60546875, + "learning_rate": 3.655083752379943e-05, + "loss": 2.3476, "step": 1922 }, { - "epoch": 0.6158526821457165, - "grad_norm": 0.59375, - "learning_rate": 1.9304658303598648e-05, - "loss": 2.425, + "epoch": 0.7471406716689736, + "grad_norm": 0.609375, + "learning_rate": 3.644600481904947e-05, + "loss": 2.4215, "step": 1923 }, { - "epoch": 0.6161729383506805, - "grad_norm": 0.54296875, - "learning_rate": 1.927743702538017e-05, - "loss": 1.8974, + "epoch": 0.7475292003593891, + "grad_norm": 0.62890625, + "learning_rate": 3.634128915597895e-05, + "loss": 2.3943, "step": 1924 }, { - "epoch": 0.6164931945556446, - "grad_norm": 0.36328125, - "learning_rate": 1.925022290508108e-05, - "loss": 1.9298, + "epoch": 0.7479177290498045, + "grad_norm": 0.6328125, + "learning_rate": 3.623669072743321e-05, + "loss": 2.4181, "step": 1925 }, { - "epoch": 0.6168134507606085, - "grad_norm": 4.03125, - "learning_rate": 1.922301597674144e-05, - "loss": 2.4168, + "epoch": 0.74830625774022, + "grad_norm": 0.66796875, + "learning_rate": 3.613220972604182e-05, + "loss": 2.3593, "step": 1926 }, { - "epoch": 0.6171337069655725, - "grad_norm": 0.4296875, - "learning_rate": 1.919581627439232e-05, - "loss": 2.331, + "epoch": 0.7486947864306355, + "grad_norm": 0.6953125, + "learning_rate": 3.602784634421804e-05, + "loss": 2.3978, "step": 1927 }, { - "epoch": 0.6174539631705365, - "grad_norm": 0.54296875, - "learning_rate": 1.9168623832055786e-05, - "loss": 1.7454, + "epoch": 0.749083315121051, + "grad_norm": 0.60546875, + "learning_rate": 3.5923600774158525e-05, + "loss": 2.3903, "step": 1928 }, { - "epoch": 0.6177742193755004, - "grad_norm": 0.79296875, - "learning_rate": 1.9141438683744785e-05, - "loss": 2.4372, + "epoch": 0.7494718438114665, + "grad_norm": 0.609375, + "learning_rate": 3.581947320784299e-05, + "loss": 2.3521, "step": 1929 }, { - "epoch": 0.6180944755804644, - "grad_norm": 0.5390625, - "learning_rate": 1.9114260863463145e-05, - "loss": 2.159, + "epoch": 0.749860372501882, + "grad_norm": 0.671875, + "learning_rate": 3.571546383703379e-05, + "loss": 2.3271, "step": 1930 }, { - "epoch": 0.6184147317854284, - "grad_norm": 0.48828125, - "learning_rate": 1.908709040520556e-05, - "loss": 2.2052, + "epoch": 0.7502489011922974, + "grad_norm": 0.6796875, + "learning_rate": 3.561157285327564e-05, + "loss": 2.3692, "step": 1931 }, { - "epoch": 0.6187349879903923, - "grad_norm": 1.7734375, - "learning_rate": 1.9059927342957468e-05, - "loss": 2.6305, + "epoch": 0.7506374298827129, + "grad_norm": 0.65234375, + "learning_rate": 3.550780044789525e-05, + "loss": 2.3958, "step": 1932 }, { - "epoch": 0.6190552441953563, - "grad_norm": 0.625, - "learning_rate": 1.9032771710695106e-05, - "loss": 2.5403, + "epoch": 0.7510259585731284, + "grad_norm": 0.64453125, + "learning_rate": 3.540414681200089e-05, + "loss": 2.3941, "step": 1933 }, { - "epoch": 0.6193755004003203, - "grad_norm": 0.578125, - "learning_rate": 1.9005623542385377e-05, - "loss": 1.5766, + "epoch": 0.7514144872635439, + "grad_norm": 0.62890625, + "learning_rate": 3.5300612136482146e-05, + "loss": 2.3173, "step": 1934 }, { - "epoch": 0.6196957566052842, - "grad_norm": 0.46875, - "learning_rate": 1.8978482871985852e-05, - "loss": 2.1659, + "epoch": 0.7518030159539594, + "grad_norm": 0.68359375, + "learning_rate": 3.519719661200954e-05, + "loss": 2.4301, "step": 1935 }, { - "epoch": 0.6200160128102482, - "grad_norm": 0.6015625, - "learning_rate": 1.895134973344476e-05, - "loss": 1.9693, + "epoch": 0.7521915446443749, + "grad_norm": 0.58984375, + "learning_rate": 3.5093900429034134e-05, + "loss": 2.3899, "step": 1936 }, { - "epoch": 0.6203362690152122, - "grad_norm": 1.2265625, - "learning_rate": 1.8924224160700866e-05, - "loss": 2.5216, + "epoch": 0.7525800733347903, + "grad_norm": 0.6171875, + "learning_rate": 3.499072377778724e-05, + "loss": 2.3445, "step": 1937 }, { - "epoch": 0.6206565252201761, - "grad_norm": 0.404296875, - "learning_rate": 1.8897106187683482e-05, - "loss": 1.6666, + "epoch": 0.7529686020252058, + "grad_norm": 0.609375, + "learning_rate": 3.488766684828e-05, + "loss": 2.2973, "step": 1938 }, { - "epoch": 0.6209767814251401, - "grad_norm": 0.59375, - "learning_rate": 1.8869995848312445e-05, - "loss": 2.3252, + "epoch": 0.7533571307156213, + "grad_norm": 0.63671875, + "learning_rate": 3.4784729830303096e-05, + "loss": 2.3857, "step": 1939 }, { - "epoch": 0.6212970376301041, - "grad_norm": 0.5, - "learning_rate": 1.8842893176498003e-05, - "loss": 2.1655, + "epoch": 0.7537456594060368, + "grad_norm": 0.671875, + "learning_rate": 3.468191291342644e-05, + "loss": 2.3077, "step": 1940 }, { - "epoch": 0.621617293835068, - "grad_norm": 0.67578125, - "learning_rate": 1.881579820614082e-05, - "loss": 2.7821, + "epoch": 0.7541341880964523, + "grad_norm": 0.625, + "learning_rate": 3.4579216286998615e-05, + "loss": 2.3862, "step": 1941 }, { - "epoch": 0.621937550040032, - "grad_norm": 0.56640625, - "learning_rate": 1.878871097113196e-05, - "loss": 2.2319, + "epoch": 0.7545227167868678, + "grad_norm": 0.62109375, + "learning_rate": 3.4476640140146796e-05, + "loss": 2.3252, "step": 1942 }, { - "epoch": 0.622257806244996, - "grad_norm": 0.9296875, - "learning_rate": 1.876163150535277e-05, - "loss": 2.2277, + "epoch": 0.7549112454772832, + "grad_norm": 0.6328125, + "learning_rate": 3.437418466177631e-05, + "loss": 2.3137, "step": 1943 }, { - "epoch": 0.62257806244996, - "grad_norm": 0.51953125, - "learning_rate": 1.8734559842674908e-05, - "loss": 1.8544, + "epoch": 0.7552997741676987, + "grad_norm": 0.6171875, + "learning_rate": 3.427185004057011e-05, + "loss": 2.3269, "step": 1944 }, { - "epoch": 0.622898318654924, - "grad_norm": 0.50390625, - "learning_rate": 1.8707496016960262e-05, - "loss": 2.4545, + "epoch": 0.7556883028581142, + "grad_norm": 0.625, + "learning_rate": 3.4169636464988695e-05, + "loss": 2.3205, "step": 1945 }, { - "epoch": 0.623218574859888, - "grad_norm": 0.73046875, - "learning_rate": 1.8680440062060895e-05, - "loss": 2.2949, + "epoch": 0.7560768315485297, + "grad_norm": 0.6640625, + "learning_rate": 3.4067544123269646e-05, + "loss": 2.3266, "step": 1946 }, { - "epoch": 0.6235388310648519, - "grad_norm": 0.474609375, - "learning_rate": 1.8653392011819072e-05, - "loss": 2.4566, + "epoch": 0.7564653602389452, + "grad_norm": 0.62109375, + "learning_rate": 3.396557320342724e-05, + "loss": 2.4334, "step": 1947 }, { - "epoch": 0.6238590872698159, - "grad_norm": 0.294921875, - "learning_rate": 1.8626351900067135e-05, - "loss": 1.9222, + "epoch": 0.7568538889293607, + "grad_norm": 0.609375, + "learning_rate": 3.386372389325213e-05, + "loss": 2.3688, "step": 1948 }, { - "epoch": 0.6241793434747798, - "grad_norm": 0.462890625, - "learning_rate": 1.8599319760627494e-05, - "loss": 2.7594, + "epoch": 0.7572424176197761, + "grad_norm": 0.62109375, + "learning_rate": 3.3761996380311067e-05, + "loss": 2.3101, "step": 1949 }, { - "epoch": 0.6244995996797438, - "grad_norm": 0.6875, - "learning_rate": 1.8572295627312616e-05, - "loss": 2.2509, + "epoch": 0.7576309463101916, + "grad_norm": 0.65625, + "learning_rate": 3.3660390851946456e-05, + "loss": 2.381, "step": 1950 }, { - "epoch": 0.6248198558847078, - "grad_norm": 0.51953125, - "learning_rate": 1.8545279533924934e-05, - "loss": 2.6185, + "epoch": 0.7580194750006071, + "grad_norm": 0.609375, + "learning_rate": 3.355890749527608e-05, + "loss": 2.3385, "step": 1951 }, { - "epoch": 0.6251401120896717, - "grad_norm": 0.5390625, - "learning_rate": 1.8518271514256812e-05, - "loss": 2.1968, + "epoch": 0.7584080036910226, + "grad_norm": 0.6328125, + "learning_rate": 3.345754649719267e-05, + "loss": 2.4517, "step": 1952 }, { - "epoch": 0.6254603682946357, - "grad_norm": 1.640625, - "learning_rate": 1.8491271602090552e-05, - "loss": 2.1392, + "epoch": 0.7587965323814381, + "grad_norm": 0.6171875, + "learning_rate": 3.335630804436368e-05, + "loss": 2.3311, "step": 1953 }, { - "epoch": 0.6257806244995997, - "grad_norm": 0.54296875, - "learning_rate": 1.8464279831198288e-05, - "loss": 2.3392, + "epoch": 0.7591850610718536, + "grad_norm": 0.62890625, + "learning_rate": 3.325519232323089e-05, + "loss": 2.3541, "step": 1954 }, { - "epoch": 0.6261008807045636, - "grad_norm": 0.89453125, - "learning_rate": 1.8437296235341954e-05, - "loss": 2.3781, + "epoch": 0.759573589762269, + "grad_norm": 0.640625, + "learning_rate": 3.315419952001001e-05, + "loss": 2.3937, "step": 1955 }, { - "epoch": 0.6264211369095276, - "grad_norm": 0.6328125, - "learning_rate": 1.8410320848273315e-05, - "loss": 2.1818, + "epoch": 0.7599621184526845, + "grad_norm": 0.60546875, + "learning_rate": 3.30533298206904e-05, + "loss": 2.3497, "step": 1956 }, { - "epoch": 0.6267413931144916, - "grad_norm": 0.36328125, - "learning_rate": 1.8383353703733803e-05, - "loss": 2.3265, + "epoch": 0.7603506471431, + "grad_norm": 0.62890625, + "learning_rate": 3.295258341103475e-05, + "loss": 2.3502, "step": 1957 }, { - "epoch": 0.6270616493194555, - "grad_norm": 0.67578125, - "learning_rate": 1.8356394835454596e-05, - "loss": 2.1143, + "epoch": 0.7607391758335155, + "grad_norm": 0.6484375, + "learning_rate": 3.285196047657865e-05, + "loss": 2.4155, "step": 1958 }, { - "epoch": 0.6273819055244195, - "grad_norm": 0.427734375, - "learning_rate": 1.8329444277156483e-05, - "loss": 1.686, + "epoch": 0.761127704523931, + "grad_norm": 0.62109375, + "learning_rate": 3.27514612026303e-05, + "loss": 2.3778, "step": 1959 }, { - "epoch": 0.6277021617293835, - "grad_norm": 0.474609375, - "learning_rate": 1.8302502062549858e-05, - "loss": 2.2293, + "epoch": 0.7615162332143465, + "grad_norm": 0.59765625, + "learning_rate": 3.265108577427022e-05, + "loss": 2.2699, "step": 1960 }, { - "epoch": 0.6280224179343474, - "grad_norm": 0.462890625, - "learning_rate": 1.8275568225334722e-05, - "loss": 2.3718, + "epoch": 0.7619047619047619, + "grad_norm": 0.63671875, + "learning_rate": 3.2550834376350745e-05, + "loss": 2.3854, "step": 1961 }, { - "epoch": 0.6283426741393114, - "grad_norm": 0.62890625, - "learning_rate": 1.824864279920054e-05, - "loss": 2.0935, + "epoch": 0.7622932905951774, + "grad_norm": 0.640625, + "learning_rate": 3.245070719349591e-05, + "loss": 2.3895, "step": 1962 }, { - "epoch": 0.6286629303442755, - "grad_norm": 0.48828125, - "learning_rate": 1.8221725817826286e-05, - "loss": 2.1848, + "epoch": 0.7626818192855929, + "grad_norm": 0.62109375, + "learning_rate": 3.235070441010092e-05, + "loss": 2.3569, "step": 1963 }, { - "epoch": 0.6289831865492393, - "grad_norm": 0.5625, - "learning_rate": 1.819481731488038e-05, - "loss": 2.5737, + "epoch": 0.7630703479760084, + "grad_norm": 0.609375, + "learning_rate": 3.225082621033193e-05, + "loss": 2.3149, "step": 1964 }, { - "epoch": 0.6293034427542034, - "grad_norm": 0.59765625, - "learning_rate": 1.8167917324020615e-05, - "loss": 2.2572, + "epoch": 0.7634588766664239, + "grad_norm": 0.6171875, + "learning_rate": 3.215107277812563e-05, + "loss": 2.3813, "step": 1965 }, { - "epoch": 0.6296236989591674, - "grad_norm": 0.466796875, - "learning_rate": 1.814102587889414e-05, - "loss": 2.2202, + "epoch": 0.7638474053568394, + "grad_norm": 0.65234375, + "learning_rate": 3.2051444297188973e-05, + "loss": 2.4437, "step": 1966 }, { - "epoch": 0.6299439551641313, - "grad_norm": 1.921875, - "learning_rate": 1.8114143013137434e-05, - "loss": 2.5762, + "epoch": 0.7642359340472548, + "grad_norm": 0.66015625, + "learning_rate": 3.195194095099874e-05, + "loss": 2.3745, "step": 1967 }, { - "epoch": 0.6302642113690953, - "grad_norm": 0.50390625, - "learning_rate": 1.8087268760376237e-05, - "loss": 2.2381, + "epoch": 0.7646244627376703, + "grad_norm": 0.62890625, + "learning_rate": 3.185256292280134e-05, + "loss": 2.415, "step": 1968 }, { - "epoch": 0.6305844675740593, - "grad_norm": 0.5390625, - "learning_rate": 1.806040315422548e-05, - "loss": 2.179, + "epoch": 0.7650129914280858, + "grad_norm": 0.61328125, + "learning_rate": 3.1753310395612315e-05, + "loss": 2.327, "step": 1969 }, { - "epoch": 0.6309047237790232, - "grad_norm": 0.5625, - "learning_rate": 1.8033546228289347e-05, - "loss": 2.1213, + "epoch": 0.7654015201185013, + "grad_norm": 0.671875, + "learning_rate": 3.165418355221612e-05, + "loss": 2.2945, "step": 1970 }, { - "epoch": 0.6312249799839872, - "grad_norm": 0.462890625, - "learning_rate": 1.8006698016161097e-05, - "loss": 1.9795, + "epoch": 0.7657900488089168, + "grad_norm": 0.59765625, + "learning_rate": 3.155518257516577e-05, + "loss": 2.3694, "step": 1971 }, { - "epoch": 0.6315452361889512, - "grad_norm": 0.5234375, - "learning_rate": 1.7979858551423147e-05, - "loss": 2.1043, + "epoch": 0.7661785774993323, + "grad_norm": 0.625, + "learning_rate": 3.145630764678247e-05, + "loss": 2.3799, "step": 1972 }, { - "epoch": 0.6318654923939151, - "grad_norm": 0.546875, - "learning_rate": 1.795302786764692e-05, - "loss": 1.9355, + "epoch": 0.7665671061897477, + "grad_norm": 0.65625, + "learning_rate": 3.135755894915526e-05, + "loss": 2.3418, "step": 1973 }, { - "epoch": 0.6321857485988791, - "grad_norm": 0.55859375, - "learning_rate": 1.7926205998392888e-05, - "loss": 2.4791, + "epoch": 0.7669556348801632, + "grad_norm": 0.609375, + "learning_rate": 3.125893666414077e-05, + "loss": 2.3477, "step": 1974 }, { - "epoch": 0.6325060048038431, - "grad_norm": 0.482421875, - "learning_rate": 1.7899392977210498e-05, - "loss": 2.5029, + "epoch": 0.7673441635705787, + "grad_norm": 0.67578125, + "learning_rate": 3.116044097336277e-05, + "loss": 2.3125, "step": 1975 }, { - "epoch": 0.632826261008807, - "grad_norm": 0.79296875, - "learning_rate": 1.787258883763812e-05, - "loss": 2.1131, + "epoch": 0.7677326922609942, + "grad_norm": 0.59375, + "learning_rate": 3.1062072058211965e-05, + "loss": 2.4135, "step": 1976 }, { - "epoch": 0.633146517213771, - "grad_norm": 0.451171875, - "learning_rate": 1.7845793613203e-05, - "loss": 2.3818, + "epoch": 0.7681212209514097, + "grad_norm": 0.6171875, + "learning_rate": 3.0963830099845484e-05, + "loss": 2.3376, "step": 1977 }, { - "epoch": 0.633466773418735, - "grad_norm": 0.41015625, - "learning_rate": 1.7819007337421272e-05, - "loss": 2.3837, + "epoch": 0.7685097496418252, + "grad_norm": 0.61328125, + "learning_rate": 3.086571527918673e-05, + "loss": 2.2661, "step": 1978 }, { - "epoch": 0.6337870296236989, - "grad_norm": 0.9140625, - "learning_rate": 1.7792230043797845e-05, - "loss": 2.3031, + "epoch": 0.7688982783322406, + "grad_norm": 0.59765625, + "learning_rate": 3.0767727776924946e-05, + "loss": 2.3262, "step": 1979 }, { - "epoch": 0.6341072858286629, - "grad_norm": 0.404296875, - "learning_rate": 1.776546176582639e-05, - "loss": 2.1035, + "epoch": 0.7692868070226561, + "grad_norm": 0.62109375, + "learning_rate": 3.066986777351492e-05, + "loss": 2.3649, "step": 1980 }, { - "epoch": 0.634427542033627, - "grad_norm": 0.5390625, - "learning_rate": 1.773870253698933e-05, - "loss": 2.4333, + "epoch": 0.7696753357130716, + "grad_norm": 0.6328125, + "learning_rate": 3.0572135449176606e-05, + "loss": 2.3777, "step": 1981 }, { - "epoch": 0.6347477982385908, - "grad_norm": 0.421875, - "learning_rate": 1.7711952390757736e-05, - "loss": 2.0962, + "epoch": 0.7700638644034871, + "grad_norm": 0.59375, + "learning_rate": 3.0474530983894857e-05, + "loss": 2.3639, "step": 1982 }, { - "epoch": 0.6350680544435549, - "grad_norm": 0.8046875, - "learning_rate": 1.7685211360591353e-05, - "loss": 2.0404, + "epoch": 0.7704523930939026, + "grad_norm": 0.66015625, + "learning_rate": 3.037705455741903e-05, + "loss": 2.3889, "step": 1983 }, { - "epoch": 0.6353883106485189, - "grad_norm": 0.46875, - "learning_rate": 1.7658479479938493e-05, - "loss": 1.8208, + "epoch": 0.770840921784318, + "grad_norm": 0.6328125, + "learning_rate": 3.0279706349262727e-05, + "loss": 2.4356, "step": 1984 }, { - "epoch": 0.6357085668534828, - "grad_norm": 0.58203125, - "learning_rate": 1.7631756782236024e-05, - "loss": 1.9733, + "epoch": 0.7712294504747335, + "grad_norm": 0.62109375, + "learning_rate": 3.0182486538703325e-05, + "loss": 2.3708, "step": 1985 }, { - "epoch": 0.6360288230584468, - "grad_norm": 0.484375, - "learning_rate": 1.760504330090937e-05, - "loss": 2.1959, + "epoch": 0.771617979165149, + "grad_norm": 0.609375, + "learning_rate": 3.008539530478186e-05, + "loss": 2.3059, "step": 1986 }, { - "epoch": 0.6363490792634108, - "grad_norm": 0.7421875, - "learning_rate": 1.757833906937237e-05, - "loss": 2.3309, + "epoch": 0.7720065078555645, + "grad_norm": 0.60546875, + "learning_rate": 2.9988432826302537e-05, + "loss": 2.365, "step": 1987 }, { - "epoch": 0.6366693354683747, - "grad_norm": 0.486328125, - "learning_rate": 1.7551644121027317e-05, - "loss": 2.45, + "epoch": 0.77239503654598, + "grad_norm": 0.625, + "learning_rate": 2.9891599281832396e-05, + "loss": 2.3592, "step": 1988 }, { - "epoch": 0.6369895916733387, - "grad_norm": 0.625, - "learning_rate": 1.7524958489264913e-05, - "loss": 2.2078, + "epoch": 0.7727835652363955, + "grad_norm": 0.59765625, + "learning_rate": 2.9794894849701083e-05, + "loss": 2.3562, "step": 1989 }, { - "epoch": 0.6373098478783027, - "grad_norm": 0.56640625, - "learning_rate": 1.7498282207464185e-05, - "loss": 2.45, + "epoch": 0.7731720939268109, + "grad_norm": 0.6796875, + "learning_rate": 2.9698319708000467e-05, + "loss": 2.3798, "step": 1990 }, { - "epoch": 0.6376301040832666, - "grad_norm": 0.44921875, - "learning_rate": 1.7471615308992444e-05, - "loss": 2.5175, + "epoch": 0.7735606226172264, + "grad_norm": 0.640625, + "learning_rate": 2.96018740345843e-05, + "loss": 2.3196, "step": 1991 }, { - "epoch": 0.6379503602882306, - "grad_norm": 0.84765625, - "learning_rate": 1.7444957827205318e-05, - "loss": 1.9297, + "epoch": 0.7739491513076419, + "grad_norm": 0.6171875, + "learning_rate": 2.9505558007067903e-05, + "loss": 2.2949, "step": 1992 }, { - "epoch": 0.6382706164931946, - "grad_norm": 0.546875, - "learning_rate": 1.741830979544661e-05, - "loss": 2.3512, + "epoch": 0.7743376799980574, + "grad_norm": 0.6328125, + "learning_rate": 2.9409371802827857e-05, + "loss": 2.3301, "step": 1993 }, { - "epoch": 0.6385908726981585, - "grad_norm": 0.546875, - "learning_rate": 1.7391671247048315e-05, - "loss": 2.4583, + "epoch": 0.7747262086884729, + "grad_norm": 0.6171875, + "learning_rate": 2.9313315599001633e-05, + "loss": 2.3685, "step": 1994 }, { - "epoch": 0.6389111289031225, - "grad_norm": 0.48828125, - "learning_rate": 1.7365042215330584e-05, - "loss": 2.1487, + "epoch": 0.7751147373788884, + "grad_norm": 0.59375, + "learning_rate": 2.9217389572487352e-05, + "loss": 2.362, "step": 1995 }, { - "epoch": 0.6392313851080864, - "grad_norm": 0.455078125, - "learning_rate": 1.7338422733601636e-05, - "loss": 1.959, + "epoch": 0.7755032660693038, + "grad_norm": 0.7265625, + "learning_rate": 2.9121593899943277e-05, + "loss": 2.3637, "step": 1996 }, { - "epoch": 0.6395516413130504, - "grad_norm": 0.484375, - "learning_rate": 1.7311812835157774e-05, - "loss": 2.6023, + "epoch": 0.7758917947597193, + "grad_norm": 0.59375, + "learning_rate": 2.9025928757787736e-05, + "loss": 2.325, "step": 1997 }, { - "epoch": 0.6398718975180144, - "grad_norm": 0.546875, - "learning_rate": 1.7285212553283295e-05, - "loss": 2.2752, + "epoch": 0.7762803234501348, + "grad_norm": 0.59375, + "learning_rate": 2.8930394322198608e-05, + "loss": 2.272, "step": 1998 }, { - "epoch": 0.6401921537229783, - "grad_norm": 0.515625, - "learning_rate": 1.7258621921250454e-05, - "loss": 1.882, + "epoch": 0.7766688521405503, + "grad_norm": 0.6484375, + "learning_rate": 2.883499076911307e-05, + "loss": 2.3975, "step": 1999 }, { - "epoch": 0.6405124099279423, - "grad_norm": 2.875, - "learning_rate": 1.7232040972319474e-05, - "loss": 2.0141, + "epoch": 0.7770573808309658, + "grad_norm": 0.65625, + "learning_rate": 2.8739718274227277e-05, + "loss": 2.3345, "step": 2000 }, { - "epoch": 0.6408326661329063, - "grad_norm": 0.375, - "learning_rate": 1.7205469739738442e-05, - "loss": 1.99, + "epoch": 0.7774459095213813, + "grad_norm": 0.61328125, + "learning_rate": 2.864457701299602e-05, + "loss": 2.3376, "step": 2001 }, { - "epoch": 0.6411529223378702, - "grad_norm": 0.31640625, - "learning_rate": 1.7178908256743277e-05, - "loss": 1.6528, + "epoch": 0.7778344382117967, + "grad_norm": 0.65234375, + "learning_rate": 2.8549567160632397e-05, + "loss": 2.3642, "step": 2002 }, { - "epoch": 0.6414731785428343, - "grad_norm": 0.91796875, - "learning_rate": 1.7152356556557745e-05, - "loss": 2.0187, + "epoch": 0.7782229669022122, + "grad_norm": 0.625, + "learning_rate": 2.845468889210752e-05, + "loss": 2.3518, "step": 2003 }, { - "epoch": 0.6417934347477983, - "grad_norm": 0.31640625, - "learning_rate": 1.7125814672393337e-05, - "loss": 2.4054, + "epoch": 0.7786114955926277, + "grad_norm": 0.61328125, + "learning_rate": 2.8359942382150194e-05, + "loss": 2.3389, "step": 2004 }, { - "epoch": 0.6421136909527622, - "grad_norm": 0.65625, - "learning_rate": 1.7099282637449273e-05, - "loss": 1.9533, + "epoch": 0.7790000242830432, + "grad_norm": 0.640625, + "learning_rate": 2.826532780524649e-05, + "loss": 2.3791, "step": 2005 }, { - "epoch": 0.6424339471577262, - "grad_norm": 0.515625, - "learning_rate": 1.7072760484912482e-05, - "loss": 2.1352, + "epoch": 0.7793885529734587, + "grad_norm": 0.60546875, + "learning_rate": 2.8170845335639595e-05, + "loss": 2.3456, "step": 2006 }, { - "epoch": 0.6427542033626902, - "grad_norm": 0.640625, - "learning_rate": 1.70462482479575e-05, - "loss": 2.5819, + "epoch": 0.7797770816638742, + "grad_norm": 0.62109375, + "learning_rate": 2.8076495147329375e-05, + "loss": 2.3348, "step": 2007 }, { - "epoch": 0.6430744595676541, - "grad_norm": 0.48828125, - "learning_rate": 1.701974595974645e-05, - "loss": 2.4939, + "epoch": 0.7801656103542896, + "grad_norm": 0.6171875, + "learning_rate": 2.79822774140721e-05, + "loss": 2.3734, "step": 2008 }, { - "epoch": 0.6433947157726181, - "grad_norm": 0.578125, - "learning_rate": 1.6993253653429063e-05, - "loss": 2.0688, + "epoch": 0.780554139044705, + "grad_norm": 0.5859375, + "learning_rate": 2.7888192309380102e-05, + "loss": 2.3042, "step": 2009 }, { - "epoch": 0.6437149719775821, - "grad_norm": 0.5390625, - "learning_rate": 1.6966771362142532e-05, - "loss": 2.3022, + "epoch": 0.7809426677351206, + "grad_norm": 0.62109375, + "learning_rate": 2.7794240006521444e-05, + "loss": 2.3699, "step": 2010 }, { - "epoch": 0.644035228182546, - "grad_norm": 0.466796875, - "learning_rate": 1.694029911901156e-05, - "loss": 2.4128, + "epoch": 0.7813311964255361, + "grad_norm": 0.640625, + "learning_rate": 2.7700420678519647e-05, + "loss": 2.3301, "step": 2011 }, { - "epoch": 0.64435548438751, - "grad_norm": 0.453125, - "learning_rate": 1.691383695714826e-05, - "loss": 1.9259, + "epoch": 0.7817197251159516, + "grad_norm": 0.58984375, + "learning_rate": 2.7606734498153365e-05, + "loss": 2.3618, "step": 2012 }, { - "epoch": 0.644675740592474, - "grad_norm": 0.63671875, - "learning_rate": 1.688738490965212e-05, - "loss": 2.2305, + "epoch": 0.7821082538063671, + "grad_norm": 0.65234375, + "learning_rate": 2.7513181637955943e-05, + "loss": 2.3884, "step": 2013 }, { - "epoch": 0.6449959967974379, - "grad_norm": 0.45703125, - "learning_rate": 1.686094300961003e-05, - "loss": 1.9657, + "epoch": 0.7824967824967825, + "grad_norm": 0.61328125, + "learning_rate": 2.7419762270215312e-05, + "loss": 2.3116, "step": 2014 }, { - "epoch": 0.6453162530024019, - "grad_norm": 0.71875, - "learning_rate": 1.683451129009614e-05, - "loss": 2.2175, + "epoch": 0.782885311187198, + "grad_norm": 0.6171875, + "learning_rate": 2.7326476566973512e-05, + "loss": 2.3987, "step": 2015 }, { - "epoch": 0.6456365092073659, - "grad_norm": 0.71484375, - "learning_rate": 1.6808089784171863e-05, - "loss": 2.3302, + "epoch": 0.7832738398776135, + "grad_norm": 0.625, + "learning_rate": 2.7233324700026464e-05, + "loss": 2.2948, "step": 2016 }, { - "epoch": 0.6459567654123298, - "grad_norm": 0.52734375, - "learning_rate": 1.6781678524885882e-05, - "loss": 1.7276, + "epoch": 0.783662368568029, + "grad_norm": 0.60546875, + "learning_rate": 2.7140306840923558e-05, + "loss": 2.2709, "step": 2017 }, { - "epoch": 0.6462770216172938, - "grad_norm": 0.3984375, - "learning_rate": 1.6755277545274023e-05, - "loss": 1.9889, + "epoch": 0.7840508972584445, + "grad_norm": 0.625, + "learning_rate": 2.704742316096743e-05, + "loss": 2.4295, "step": 2018 }, { - "epoch": 0.6465972778222578, - "grad_norm": 0.408203125, - "learning_rate": 1.6728886878359258e-05, - "loss": 2.0543, + "epoch": 0.78443942594886, + "grad_norm": 0.62109375, + "learning_rate": 2.6954673831213605e-05, + "loss": 2.3513, "step": 2019 }, { - "epoch": 0.6469175340272217, - "grad_norm": 0.427734375, - "learning_rate": 1.670250655715168e-05, - "loss": 2.6313, + "epoch": 0.7848279546392753, + "grad_norm": 0.6328125, + "learning_rate": 2.6862059022470198e-05, + "loss": 2.3083, "step": 2020 }, { - "epoch": 0.6472377902321858, - "grad_norm": 0.609375, - "learning_rate": 1.667613661464842e-05, - "loss": 2.2489, + "epoch": 0.7852164833296909, + "grad_norm": 0.6328125, + "learning_rate": 2.6769578905297588e-05, + "loss": 2.3593, "step": 2021 }, { - "epoch": 0.6475580464371498, - "grad_norm": 1.515625, - "learning_rate": 1.664977708383365e-05, - "loss": 2.7663, + "epoch": 0.7856050120201064, + "grad_norm": 0.62890625, + "learning_rate": 2.667723365000804e-05, + "loss": 2.2834, "step": 2022 }, { - "epoch": 0.6478783026421137, - "grad_norm": 0.55859375, - "learning_rate": 1.6623427997678497e-05, - "loss": 2.5957, + "epoch": 0.7859935407105219, + "grad_norm": 0.73046875, + "learning_rate": 2.6585023426665535e-05, + "loss": 2.3171, "step": 2023 }, { - "epoch": 0.6481985588470777, - "grad_norm": 0.65234375, - "learning_rate": 1.659708938914102e-05, - "loss": 2.0438, + "epoch": 0.7863820694009374, + "grad_norm": 0.6015625, + "learning_rate": 2.6492948405085348e-05, + "loss": 2.2423, "step": 2024 }, { - "epoch": 0.6485188150520417, - "grad_norm": 0.4765625, - "learning_rate": 1.6570761291166207e-05, - "loss": 2.1817, + "epoch": 0.7867705980913529, + "grad_norm": 0.62890625, + "learning_rate": 2.640100875483379e-05, + "loss": 2.3983, "step": 2025 }, { - "epoch": 0.6488390712570056, - "grad_norm": 0.478515625, - "learning_rate": 1.654444373668586e-05, - "loss": 2.3496, + "epoch": 0.7871591267817682, + "grad_norm": 0.64453125, + "learning_rate": 2.6309204645227825e-05, + "loss": 2.3511, "step": 2026 }, { - "epoch": 0.6491593274619696, - "grad_norm": 0.546875, - "learning_rate": 1.6518136758618607e-05, - "loss": 2.2006, + "epoch": 0.7875476554721837, + "grad_norm": 0.6171875, + "learning_rate": 2.621753624533484e-05, + "loss": 2.2993, "step": 2027 }, { - "epoch": 0.6494795836669336, - "grad_norm": 0.5703125, - "learning_rate": 1.649184038986986e-05, - "loss": 1.8331, + "epoch": 0.7879361841625993, + "grad_norm": 0.58984375, + "learning_rate": 2.6126003723972326e-05, + "loss": 2.3746, "step": 2028 }, { - "epoch": 0.6497998398718975, - "grad_norm": 0.51171875, - "learning_rate": 1.6465554663331746e-05, - "loss": 2.3697, + "epoch": 0.7883247128530148, + "grad_norm": 0.6015625, + "learning_rate": 2.603460724970741e-05, + "loss": 2.4112, "step": 2029 }, { - "epoch": 0.6501200960768615, - "grad_norm": 0.498046875, - "learning_rate": 1.6439279611883075e-05, - "loss": 2.2204, + "epoch": 0.7887132415434303, + "grad_norm": 0.6796875, + "learning_rate": 2.594334699085682e-05, + "loss": 2.4033, "step": 2030 }, { - "epoch": 0.6504403522818255, - "grad_norm": 1.3671875, - "learning_rate": 1.6413015268389323e-05, - "loss": 2.3094, + "epoch": 0.7891017702338458, + "grad_norm": 0.6015625, + "learning_rate": 2.5852223115486385e-05, + "loss": 2.2852, "step": 2031 }, { - "epoch": 0.6507606084867894, - "grad_norm": 0.97265625, - "learning_rate": 1.6386761665702564e-05, - "loss": 2.6854, + "epoch": 0.7894902989242611, + "grad_norm": 0.61328125, + "learning_rate": 2.5761235791410698e-05, + "loss": 2.3602, "step": 2032 }, { - "epoch": 0.6510808646917534, - "grad_norm": 0.498046875, - "learning_rate": 1.6360518836661418e-05, - "loss": 2.6873, + "epoch": 0.7898788276146766, + "grad_norm": 0.60546875, + "learning_rate": 2.567038518619297e-05, + "loss": 2.3884, "step": 2033 }, { - "epoch": 0.6514011208967174, - "grad_norm": 0.419921875, - "learning_rate": 1.6334286814091072e-05, - "loss": 2.2687, + "epoch": 0.7902673563050922, + "grad_norm": 0.64453125, + "learning_rate": 2.5579671467144585e-05, + "loss": 2.4177, "step": 2034 }, { - "epoch": 0.6517213771016813, - "grad_norm": 0.337890625, - "learning_rate": 1.6308065630803144e-05, - "loss": 2.0965, + "epoch": 0.7906558849955077, + "grad_norm": 0.625, + "learning_rate": 2.5489094801324854e-05, + "loss": 2.3687, "step": 2035 }, { - "epoch": 0.6520416333066453, - "grad_norm": 0.59375, - "learning_rate": 1.6281855319595747e-05, - "loss": 2.6411, + "epoch": 0.7910444136859232, + "grad_norm": 0.60546875, + "learning_rate": 2.5398655355540658e-05, + "loss": 2.3799, "step": 2036 }, { - "epoch": 0.6523618895116093, - "grad_norm": 0.64453125, - "learning_rate": 1.6255655913253357e-05, - "loss": 2.0638, + "epoch": 0.7914329423763387, + "grad_norm": 0.59375, + "learning_rate": 2.530835329634622e-05, + "loss": 2.3293, "step": 2037 }, { - "epoch": 0.6526821457165732, - "grad_norm": 0.546875, - "learning_rate": 1.622946744454681e-05, - "loss": 2.3375, + "epoch": 0.791821471066754, + "grad_norm": 0.60546875, + "learning_rate": 2.5218188790042706e-05, + "loss": 2.3401, "step": 2038 }, { - "epoch": 0.6530024019215372, - "grad_norm": 0.71875, - "learning_rate": 1.620328994623329e-05, - "loss": 2.0346, + "epoch": 0.7922099997571695, + "grad_norm": 0.609375, + "learning_rate": 2.5128162002678024e-05, + "loss": 2.3879, "step": 2039 }, { - "epoch": 0.6533226581265013, - "grad_norm": 0.390625, - "learning_rate": 1.617712345105624e-05, - "loss": 2.1022, + "epoch": 0.792598528447585, + "grad_norm": 0.5859375, + "learning_rate": 2.503827310004635e-05, + "loss": 2.3356, "step": 2040 }, { - "epoch": 0.6536429143314652, - "grad_norm": 0.7265625, - "learning_rate": 1.6150967991745323e-05, - "loss": 2.2443, + "epoch": 0.7929870571380006, + "grad_norm": 0.625, + "learning_rate": 2.4948522247688023e-05, + "loss": 2.3512, "step": 2041 }, { - "epoch": 0.6539631705364292, - "grad_norm": 0.515625, - "learning_rate": 1.6124823601016437e-05, - "loss": 1.8261, + "epoch": 0.793375585828416, + "grad_norm": 1.0546875, + "learning_rate": 2.4858909610889114e-05, + "loss": 2.3043, "step": 2042 }, { - "epoch": 0.6542834267413931, - "grad_norm": 0.58203125, - "learning_rate": 1.6098690311571608e-05, - "loss": 1.3938, + "epoch": 0.7937641145188316, + "grad_norm": 0.63671875, + "learning_rate": 2.476943535468117e-05, + "loss": 2.3064, "step": 2043 }, { - "epoch": 0.6546036829463571, - "grad_norm": 0.46484375, - "learning_rate": 1.6072568156098972e-05, - "loss": 2.2079, + "epoch": 0.794152643209247, + "grad_norm": 0.609375, + "learning_rate": 2.4680099643840882e-05, + "loss": 2.3251, "step": 2044 }, { - "epoch": 0.6549239391513211, - "grad_norm": 0.265625, - "learning_rate": 1.6046457167272773e-05, - "loss": 1.7936, + "epoch": 0.7945411718996624, + "grad_norm": 0.61328125, + "learning_rate": 2.459090264288979e-05, + "loss": 2.3401, "step": 2045 }, { - "epoch": 0.655244195356285, - "grad_norm": 0.56640625, - "learning_rate": 1.6020357377753237e-05, - "loss": 2.3265, + "epoch": 0.794929700590078, + "grad_norm": 0.62890625, + "learning_rate": 2.4501844516094007e-05, + "loss": 2.2794, "step": 2046 }, { - "epoch": 0.655564451561249, - "grad_norm": 0.51953125, - "learning_rate": 1.5994268820186624e-05, - "loss": 2.1049, + "epoch": 0.7953182292804934, + "grad_norm": 0.6171875, + "learning_rate": 2.4412925427463874e-05, + "loss": 2.3531, "step": 2047 }, { - "epoch": 0.655884707766213, - "grad_norm": 0.48046875, - "learning_rate": 1.5968191527205134e-05, - "loss": 2.2532, + "epoch": 0.795706757970909, + "grad_norm": 0.63671875, + "learning_rate": 2.4324145540753697e-05, + "loss": 2.3799, "step": 2048 }, { - "epoch": 0.6562049639711769, - "grad_norm": 0.56640625, - "learning_rate": 1.5942125531426848e-05, - "loss": 2.5682, + "epoch": 0.7960952866613243, + "grad_norm": 0.609375, + "learning_rate": 2.4235505019461367e-05, + "loss": 2.3758, "step": 2049 }, { - "epoch": 0.6565252201761409, - "grad_norm": 0.9375, - "learning_rate": 1.591607086545577e-05, - "loss": 2.6043, + "epoch": 0.7964838153517398, + "grad_norm": 0.62109375, + "learning_rate": 2.4147004026828192e-05, + "loss": 2.3625, "step": 2050 }, { - "epoch": 0.6568454763811049, - "grad_norm": 0.82421875, - "learning_rate": 1.589002756188169e-05, - "loss": 1.8284, + "epoch": 0.7968723440421553, + "grad_norm": 0.61328125, + "learning_rate": 2.4058642725838486e-05, + "loss": 2.3455, "step": 2051 }, { - "epoch": 0.6571657325860688, - "grad_norm": 0.6875, - "learning_rate": 1.5863995653280187e-05, - "loss": 2.0559, + "epoch": 0.7972608727325708, + "grad_norm": 0.6328125, + "learning_rate": 2.3970421279219323e-05, + "loss": 2.3983, "step": 2052 }, { - "epoch": 0.6574859887910328, - "grad_norm": 0.578125, - "learning_rate": 1.5837975172212615e-05, - "loss": 2.0543, + "epoch": 0.7976494014229863, + "grad_norm": 0.5859375, + "learning_rate": 2.3882339849440205e-05, + "loss": 2.3582, "step": 2053 }, { - "epoch": 0.6578062449959968, - "grad_norm": 0.57421875, - "learning_rate": 1.5811966151226014e-05, - "loss": 2.0767, + "epoch": 0.7980379301134018, + "grad_norm": 0.61328125, + "learning_rate": 2.3794398598712786e-05, + "loss": 2.3543, "step": 2054 }, { - "epoch": 0.6581265012009607, - "grad_norm": 0.94921875, - "learning_rate": 1.578596862285308e-05, - "loss": 2.3316, + "epoch": 0.7984264588038172, + "grad_norm": 0.6171875, + "learning_rate": 2.370659768899056e-05, + "loss": 2.3625, "step": 2055 }, { - "epoch": 0.6584467574059247, - "grad_norm": 0.51953125, - "learning_rate": 1.575998261961217e-05, - "loss": 2.4229, + "epoch": 0.7988149874942327, + "grad_norm": 0.64453125, + "learning_rate": 2.36189372819686e-05, + "loss": 2.3334, "step": 2056 }, { - "epoch": 0.6587670136108887, - "grad_norm": 0.4765625, - "learning_rate": 1.5734008174007182e-05, - "loss": 2.3858, + "epoch": 0.7992035161846482, + "grad_norm": 0.6328125, + "learning_rate": 2.3531417539083134e-05, + "loss": 2.429, "step": 2057 }, { - "epoch": 0.6590872698158526, - "grad_norm": 0.451171875, - "learning_rate": 1.5708045318527583e-05, - "loss": 2.5705, + "epoch": 0.7995920448750637, + "grad_norm": 0.6171875, + "learning_rate": 2.3444038621511433e-05, + "loss": 2.3277, "step": 2058 }, { - "epoch": 0.6594075260208166, - "grad_norm": 0.4140625, - "learning_rate": 1.568209408564834e-05, - "loss": 2.8622, + "epoch": 0.7999805735654792, + "grad_norm": 0.609375, + "learning_rate": 2.33568006901714e-05, + "loss": 2.3584, "step": 2059 }, { - "epoch": 0.6597277822257807, - "grad_norm": 0.48046875, - "learning_rate": 1.5656154507829867e-05, - "loss": 2.4677, + "epoch": 0.8003691022558947, + "grad_norm": 0.62109375, + "learning_rate": 2.3269703905721285e-05, + "loss": 2.4414, "step": 2060 }, { - "epoch": 0.6600480384307446, - "grad_norm": 0.478515625, - "learning_rate": 1.5630226617518033e-05, - "loss": 2.4751, + "epoch": 0.8007576309463101, + "grad_norm": 0.6328125, + "learning_rate": 2.318274842855941e-05, + "loss": 2.3893, "step": 2061 }, { - "epoch": 0.6603682946357086, - "grad_norm": 0.53515625, - "learning_rate": 1.560431044714405e-05, - "loss": 2.2698, + "epoch": 0.8011461596367256, + "grad_norm": 0.625, + "learning_rate": 2.309593441882385e-05, + "loss": 2.2949, "step": 2062 }, { - "epoch": 0.6606885508406726, - "grad_norm": 0.431640625, - "learning_rate": 1.5578406029124485e-05, - "loss": 2.0726, + "epoch": 0.8015346883271411, + "grad_norm": 0.58984375, + "learning_rate": 2.3009262036392177e-05, + "loss": 2.373, "step": 2063 }, { - "epoch": 0.6610088070456365, - "grad_norm": 0.49609375, - "learning_rate": 1.555251339586123e-05, - "loss": 1.9642, + "epoch": 0.8019232170175566, + "grad_norm": 0.578125, + "learning_rate": 2.2922731440881128e-05, + "loss": 2.392, "step": 2064 }, { - "epoch": 0.6613290632506005, - "grad_norm": 1.421875, - "learning_rate": 1.5526632579741386e-05, - "loss": 1.9934, + "epoch": 0.8023117457079721, + "grad_norm": 0.609375, + "learning_rate": 2.2836342791646325e-05, + "loss": 2.3395, "step": 2065 }, { - "epoch": 0.6616493194555645, - "grad_norm": 1.65625, - "learning_rate": 1.5500763613137316e-05, - "loss": 2.7078, + "epoch": 0.8027002743983876, + "grad_norm": 0.61328125, + "learning_rate": 2.275009624778195e-05, + "loss": 2.3918, "step": 2066 }, { - "epoch": 0.6619695756605284, - "grad_norm": 0.66796875, - "learning_rate": 1.5474906528406542e-05, - "loss": 2.3785, + "epoch": 0.803088803088803, + "grad_norm": 0.6171875, + "learning_rate": 2.2663991968120512e-05, + "loss": 2.3602, "step": 2067 }, { - "epoch": 0.6622898318654924, - "grad_norm": 0.6875, - "learning_rate": 1.544906135789172e-05, - "loss": 2.1679, + "epoch": 0.8034773317792185, + "grad_norm": 0.6328125, + "learning_rate": 2.257803011123254e-05, + "loss": 2.3232, "step": 2068 }, { - "epoch": 0.6626100880704564, - "grad_norm": 0.515625, - "learning_rate": 1.542322813392062e-05, - "loss": 2.3335, + "epoch": 0.803865860469634, + "grad_norm": 0.625, + "learning_rate": 2.2492210835426253e-05, + "loss": 2.327, "step": 2069 }, { - "epoch": 0.6629303442754203, - "grad_norm": 0.63671875, - "learning_rate": 1.5397406888806063e-05, - "loss": 2.1106, + "epoch": 0.8042543891600495, + "grad_norm": 0.61328125, + "learning_rate": 2.2406534298747293e-05, + "loss": 2.3697, "step": 2070 }, { - "epoch": 0.6632506004803843, - "grad_norm": 0.5546875, - "learning_rate": 1.537159765484587e-05, - "loss": 1.8888, + "epoch": 0.804642917850465, + "grad_norm": 0.59765625, + "learning_rate": 2.2321000658978465e-05, + "loss": 2.3967, "step": 2071 }, { - "epoch": 0.6635708566853483, - "grad_norm": 0.48046875, - "learning_rate": 1.5345800464322856e-05, - "loss": 2.4113, + "epoch": 0.8050314465408805, + "grad_norm": 0.609375, + "learning_rate": 2.2235610073639335e-05, + "loss": 2.326, "step": 2072 }, { - "epoch": 0.6638911128903122, - "grad_norm": 0.431640625, - "learning_rate": 1.5320015349504778e-05, - "loss": 1.6792, + "epoch": 0.8054199752312959, + "grad_norm": 0.6171875, + "learning_rate": 2.2150362699986083e-05, + "loss": 2.3649, "step": 2073 }, { - "epoch": 0.6642113690952762, - "grad_norm": 0.70703125, - "learning_rate": 1.529424234264426e-05, - "loss": 2.4077, + "epoch": 0.8058085039217114, + "grad_norm": 0.625, + "learning_rate": 2.206525869501114e-05, + "loss": 2.4364, "step": 2074 }, { - "epoch": 0.6645316253002402, - "grad_norm": 2.109375, - "learning_rate": 1.5268481475978817e-05, - "loss": 2.5355, + "epoch": 0.8061970326121269, + "grad_norm": 0.5859375, + "learning_rate": 2.198029821544292e-05, + "loss": 2.3134, "step": 2075 }, { - "epoch": 0.6648518815052041, - "grad_norm": 0.82421875, - "learning_rate": 1.5242732781730749e-05, - "loss": 2.0814, + "epoch": 0.8065855613025424, + "grad_norm": 0.60546875, + "learning_rate": 2.1895481417745457e-05, + "loss": 2.3823, "step": 2076 }, { - "epoch": 0.6651721377101681, - "grad_norm": 0.5703125, - "learning_rate": 1.5216996292107144e-05, - "loss": 2.5232, + "epoch": 0.8069740899929579, + "grad_norm": 0.62890625, + "learning_rate": 2.1810808458118247e-05, + "loss": 2.4092, "step": 2077 }, { - "epoch": 0.6654923939151322, - "grad_norm": 0.54296875, - "learning_rate": 1.5191272039299825e-05, - "loss": 2.3116, + "epoch": 0.8073626186833734, + "grad_norm": 0.64453125, + "learning_rate": 2.172627949249586e-05, + "loss": 2.3765, "step": 2078 }, { - "epoch": 0.665812650120096, - "grad_norm": 0.51171875, - "learning_rate": 1.5165560055485301e-05, - "loss": 1.9626, + "epoch": 0.8077511473737888, + "grad_norm": 0.609375, + "learning_rate": 2.1641894676547703e-05, + "loss": 2.3685, "step": 2079 }, { - "epoch": 0.6661329063250601, - "grad_norm": 0.349609375, - "learning_rate": 1.5139860372824741e-05, - "loss": 1.8281, + "epoch": 0.8081396760642043, + "grad_norm": 0.61328125, + "learning_rate": 2.15576541656777e-05, + "loss": 2.2613, "step": 2080 }, { - "epoch": 0.6664531625300241, - "grad_norm": 0.470703125, - "learning_rate": 1.5114173023463932e-05, - "loss": 2.2899, + "epoch": 0.8085282047546198, + "grad_norm": 0.63671875, + "learning_rate": 2.1473558115024027e-05, + "loss": 2.2979, "step": 2081 }, { - "epoch": 0.666773418734988, - "grad_norm": 0.46484375, - "learning_rate": 1.5088498039533219e-05, - "loss": 1.9186, + "epoch": 0.8089167334450353, + "grad_norm": 0.609375, + "learning_rate": 2.1389606679458828e-05, + "loss": 2.3692, "step": 2082 }, { - "epoch": 0.667093674939952, - "grad_norm": 0.48828125, - "learning_rate": 1.5062835453147495e-05, - "loss": 2.5786, + "epoch": 0.8093052621354508, + "grad_norm": 0.6015625, + "learning_rate": 2.1305800013587908e-05, + "loss": 2.3236, "step": 2083 }, { - "epoch": 0.667413931144916, - "grad_norm": 0.63671875, - "learning_rate": 1.5037185296406142e-05, - "loss": 2.3842, + "epoch": 0.8096937908258663, + "grad_norm": 0.59375, + "learning_rate": 2.122213827175048e-05, + "loss": 2.3073, "step": 2084 }, { - "epoch": 0.6677341873498799, - "grad_norm": 0.515625, - "learning_rate": 1.5011547601392983e-05, - "loss": 2.313, + "epoch": 0.8100823195162817, + "grad_norm": 0.63671875, + "learning_rate": 2.113862160801883e-05, + "loss": 2.3627, "step": 2085 }, { - "epoch": 0.6680544435548439, - "grad_norm": 0.578125, - "learning_rate": 1.4985922400176278e-05, - "loss": 2.6968, + "epoch": 0.8104708482066972, + "grad_norm": 0.60546875, + "learning_rate": 2.1055250176198094e-05, + "loss": 2.2891, "step": 2086 }, { - "epoch": 0.6683746997598079, - "grad_norm": 0.455078125, - "learning_rate": 1.4960309724808633e-05, - "loss": 1.9038, + "epoch": 0.8108593768971127, + "grad_norm": 0.59765625, + "learning_rate": 2.097202412982595e-05, + "loss": 2.3075, "step": 2087 }, { - "epoch": 0.6686949559647718, - "grad_norm": 0.8359375, - "learning_rate": 1.4934709607327001e-05, - "loss": 1.8091, + "epoch": 0.8112479055875282, + "grad_norm": 0.59375, + "learning_rate": 2.088894362217233e-05, + "loss": 2.3189, "step": 2088 }, { - "epoch": 0.6690152121697358, - "grad_norm": 0.5078125, - "learning_rate": 1.4909122079752632e-05, - "loss": 2.426, + "epoch": 0.8116364342779437, + "grad_norm": 0.62890625, + "learning_rate": 2.080600880623913e-05, + "loss": 2.2979, "step": 2089 }, { - "epoch": 0.6693354683746997, - "grad_norm": 0.4140625, - "learning_rate": 1.4883547174091011e-05, - "loss": 1.9204, + "epoch": 0.8120249629683592, + "grad_norm": 0.59375, + "learning_rate": 2.0723219834759945e-05, + "loss": 2.3131, "step": 2090 }, { - "epoch": 0.6696557245796637, - "grad_norm": 0.55078125, - "learning_rate": 1.4857984922331847e-05, - "loss": 2.1271, + "epoch": 0.8124134916587746, + "grad_norm": 0.5859375, + "learning_rate": 2.0640576860199778e-05, + "loss": 2.3436, "step": 2091 }, { - "epoch": 0.6699759807846277, - "grad_norm": 0.53515625, - "learning_rate": 1.4832435356449026e-05, - "loss": 2.5165, + "epoch": 0.8128020203491901, + "grad_norm": 0.60546875, + "learning_rate": 2.05580800347548e-05, + "loss": 2.4079, "step": 2092 }, { - "epoch": 0.6702962369895916, - "grad_norm": 0.388671875, - "learning_rate": 1.4806898508400552e-05, - "loss": 1.994, + "epoch": 0.8131905490396056, + "grad_norm": 0.6015625, + "learning_rate": 2.0475729510351937e-05, + "loss": 2.3397, "step": 2093 }, { - "epoch": 0.6706164931945556, - "grad_norm": 0.53515625, - "learning_rate": 1.478137441012853e-05, - "loss": 2.0746, + "epoch": 0.8135790777300211, + "grad_norm": 0.60546875, + "learning_rate": 2.0393525438648774e-05, + "loss": 2.4194, "step": 2094 }, { - "epoch": 0.6709367493995196, - "grad_norm": 0.373046875, - "learning_rate": 1.4755863093559118e-05, - "loss": 1.7466, + "epoch": 0.8139676064204366, + "grad_norm": 0.62109375, + "learning_rate": 2.031146797103317e-05, + "loss": 2.3762, "step": 2095 }, { - "epoch": 0.6712570056044835, - "grad_norm": 0.51953125, - "learning_rate": 1.473036459060248e-05, - "loss": 2.247, + "epoch": 0.8143561351108521, + "grad_norm": 0.5859375, + "learning_rate": 2.0229557258622977e-05, + "loss": 2.402, "step": 2096 }, { - "epoch": 0.6715772618094475, - "grad_norm": 0.71484375, - "learning_rate": 1.4704878933152761e-05, - "loss": 2.4396, + "epoch": 0.8147446638012675, + "grad_norm": 0.61328125, + "learning_rate": 2.0147793452265796e-05, + "loss": 2.3029, "step": 2097 }, { - "epoch": 0.6718975180144116, - "grad_norm": 0.5078125, - "learning_rate": 1.4679406153088032e-05, - "loss": 2.3092, + "epoch": 0.815133192491683, + "grad_norm": 0.65625, + "learning_rate": 2.0066176702538676e-05, + "loss": 2.3814, "step": 2098 }, { - "epoch": 0.6722177742193755, - "grad_norm": 0.5625, - "learning_rate": 1.4653946282270258e-05, - "loss": 2.0661, + "epoch": 0.8155217211820985, + "grad_norm": 0.6171875, + "learning_rate": 1.998470715974784e-05, + "loss": 2.3413, "step": 2099 }, { - "epoch": 0.6725380304243395, - "grad_norm": 0.8671875, - "learning_rate": 1.4628499352545266e-05, - "loss": 2.3798, + "epoch": 0.815910249872514, + "grad_norm": 0.61328125, + "learning_rate": 1.990338497392845e-05, + "loss": 2.3015, "step": 2100 }, { - "epoch": 0.6728582866293035, - "grad_norm": 0.98046875, - "learning_rate": 1.4603065395742677e-05, - "loss": 2.1861, + "epoch": 0.8162987785629295, + "grad_norm": 0.6015625, + "learning_rate": 1.9822210294844222e-05, + "loss": 2.355, "step": 2101 }, { - "epoch": 0.6731785428342674, - "grad_norm": 0.45703125, - "learning_rate": 1.4577644443675895e-05, - "loss": 1.6774, + "epoch": 0.816687307253345, + "grad_norm": 0.60546875, + "learning_rate": 1.9741183271987284e-05, + "loss": 2.4106, "step": 2102 }, { - "epoch": 0.6734987990392314, - "grad_norm": 0.494140625, - "learning_rate": 1.4552236528142078e-05, - "loss": 2.7617, + "epoch": 0.8170758359437604, + "grad_norm": 0.5859375, + "learning_rate": 1.9660304054577816e-05, + "loss": 2.2703, "step": 2103 }, { - "epoch": 0.6738190552441954, - "grad_norm": 0.515625, - "learning_rate": 1.452684168092203e-05, - "loss": 2.9759, + "epoch": 0.8174643646341759, + "grad_norm": 0.59375, + "learning_rate": 1.9579572791563804e-05, + "loss": 2.3253, "step": 2104 }, { - "epoch": 0.6741393114491593, - "grad_norm": 0.53125, - "learning_rate": 1.4501459933780249e-05, - "loss": 2.4151, + "epoch": 0.8178528933245914, + "grad_norm": 0.61328125, + "learning_rate": 1.9498989631620767e-05, + "loss": 2.3657, "step": 2105 }, { - "epoch": 0.6744595676541233, - "grad_norm": 0.318359375, - "learning_rate": 1.4476091318464836e-05, - "loss": 2.0188, + "epoch": 0.8182414220150069, + "grad_norm": 0.6171875, + "learning_rate": 1.941855472315145e-05, + "loss": 2.4544, "step": 2106 }, { - "epoch": 0.6747798238590873, - "grad_norm": 0.474609375, - "learning_rate": 1.4450735866707474e-05, - "loss": 2.5587, + "epoch": 0.8186299507054224, + "grad_norm": 0.6015625, + "learning_rate": 1.9338268214285604e-05, + "loss": 2.36, "step": 2107 }, { - "epoch": 0.6751000800640512, - "grad_norm": 0.451171875, - "learning_rate": 1.4425393610223342e-05, - "loss": 2.2319, + "epoch": 0.8190184793958379, + "grad_norm": 0.66015625, + "learning_rate": 1.925813025287968e-05, + "loss": 2.3582, "step": 2108 }, { - "epoch": 0.6754203362690152, - "grad_norm": 0.470703125, - "learning_rate": 1.4400064580711182e-05, - "loss": 2.3308, + "epoch": 0.8194070080862533, + "grad_norm": 0.6328125, + "learning_rate": 1.9178140986516568e-05, + "loss": 2.303, "step": 2109 }, { - "epoch": 0.6757405924739792, - "grad_norm": 1.46875, - "learning_rate": 1.4374748809853125e-05, - "loss": 1.9952, + "epoch": 0.8197955367766688, + "grad_norm": 0.63671875, + "learning_rate": 1.9098300562505266e-05, + "loss": 2.3696, "step": 2110 }, { - "epoch": 0.6760608486789431, - "grad_norm": 0.482421875, - "learning_rate": 1.4349446329314756e-05, - "loss": 2.4647, + "epoch": 0.8201840654670843, + "grad_norm": 0.61328125, + "learning_rate": 1.9018609127880727e-05, + "loss": 2.3677, "step": 2111 }, { - "epoch": 0.6763811048839071, - "grad_norm": 0.419921875, - "learning_rate": 1.4324157170745025e-05, - "loss": 2.0062, + "epoch": 0.8205725941574998, + "grad_norm": 0.62890625, + "learning_rate": 1.8939066829403508e-05, + "loss": 2.3268, "step": 2112 }, { - "epoch": 0.6767013610888711, - "grad_norm": 0.53515625, - "learning_rate": 1.429888136577622e-05, - "loss": 2.5416, + "epoch": 0.8209611228479153, + "grad_norm": 0.6015625, + "learning_rate": 1.8859673813559497e-05, + "loss": 2.3276, "step": 2113 }, { - "epoch": 0.677021617293835, - "grad_norm": 0.53515625, - "learning_rate": 1.4273618946023925e-05, - "loss": 2.4491, + "epoch": 0.8213496515383308, + "grad_norm": 0.6015625, + "learning_rate": 1.8780430226559686e-05, + "loss": 2.3263, "step": 2114 }, { - "epoch": 0.677341873498799, - "grad_norm": 0.40234375, - "learning_rate": 1.4248369943086998e-05, - "loss": 2.3076, + "epoch": 0.8217381802287462, + "grad_norm": 0.61328125, + "learning_rate": 1.8701336214339883e-05, + "loss": 2.3716, "step": 2115 }, { - "epoch": 0.677662129703763, - "grad_norm": 0.59765625, - "learning_rate": 1.422313438854746e-05, - "loss": 2.4278, + "epoch": 0.8221267089191617, + "grad_norm": 0.6796875, + "learning_rate": 1.8622391922560377e-05, + "loss": 2.4096, "step": 2116 }, { - "epoch": 0.677982385908727, - "grad_norm": 0.609375, - "learning_rate": 1.419791231397059e-05, - "loss": 2.1202, + "epoch": 0.8225152376095772, + "grad_norm": 0.6171875, + "learning_rate": 1.8543597496605793e-05, + "loss": 2.3107, "step": 2117 }, { - "epoch": 0.678302642113691, - "grad_norm": 0.4296875, - "learning_rate": 1.4172703750904746e-05, - "loss": 1.8748, + "epoch": 0.8229037662999927, + "grad_norm": 0.6484375, + "learning_rate": 1.8464953081584735e-05, + "loss": 2.3928, "step": 2118 }, { - "epoch": 0.678622898318655, - "grad_norm": 0.3125, - "learning_rate": 1.4147508730881403e-05, - "loss": 2.3358, + "epoch": 0.8232922949904082, + "grad_norm": 0.62109375, + "learning_rate": 1.83864588223296e-05, + "loss": 2.3962, "step": 2119 }, { - "epoch": 0.6789431545236189, - "grad_norm": 0.7890625, - "learning_rate": 1.4122327285415105e-05, - "loss": 2.2464, + "epoch": 0.8236808236808236, + "grad_norm": 0.625, + "learning_rate": 1.830811486339613e-05, + "loss": 2.3183, "step": 2120 }, { - "epoch": 0.6792634107285829, - "grad_norm": 0.5078125, - "learning_rate": 1.4097159446003417e-05, - "loss": 2.0882, + "epoch": 0.8240693523712391, + "grad_norm": 0.59375, + "learning_rate": 1.8229921349063396e-05, + "loss": 2.3754, "step": 2121 }, { - "epoch": 0.6795836669335469, - "grad_norm": 0.4453125, - "learning_rate": 1.4072005244126846e-05, - "loss": 1.9614, + "epoch": 0.8244578810616546, + "grad_norm": 0.59765625, + "learning_rate": 1.8151878423333346e-05, + "loss": 2.3284, "step": 2122 }, { - "epoch": 0.6799039231385108, - "grad_norm": 1.28125, - "learning_rate": 1.4046864711248913e-05, - "loss": 2.0258, + "epoch": 0.8248464097520701, + "grad_norm": 0.640625, + "learning_rate": 1.807398622993064e-05, + "loss": 2.3157, "step": 2123 }, { - "epoch": 0.6802241793434748, - "grad_norm": 0.515625, - "learning_rate": 1.4021737878815966e-05, - "loss": 1.9489, + "epoch": 0.8252349384424856, + "grad_norm": 0.66796875, + "learning_rate": 1.7996244912302317e-05, + "loss": 2.3063, "step": 2124 }, { - "epoch": 0.6805444355484388, - "grad_norm": 0.58984375, - "learning_rate": 1.3996624778257287e-05, - "loss": 2.5073, + "epoch": 0.8256234671329011, + "grad_norm": 0.6015625, + "learning_rate": 1.791865461361758e-05, + "loss": 2.3508, "step": 2125 }, { - "epoch": 0.6808646917534027, - "grad_norm": 0.45703125, - "learning_rate": 1.3971525440984914e-05, - "loss": 1.9678, + "epoch": 0.8260119958233165, + "grad_norm": 0.62890625, + "learning_rate": 1.7841215476767493e-05, + "loss": 2.3617, "step": 2126 }, { - "epoch": 0.6811849479583667, - "grad_norm": 0.3984375, - "learning_rate": 1.3946439898393718e-05, - "loss": 2.1684, + "epoch": 0.826400524513732, + "grad_norm": 0.58984375, + "learning_rate": 1.7763927644364765e-05, + "loss": 2.3354, "step": 2127 }, { - "epoch": 0.6815052041633307, - "grad_norm": 0.51171875, - "learning_rate": 1.3921368181861293e-05, - "loss": 2.5053, + "epoch": 0.8267890532041475, + "grad_norm": 0.58203125, + "learning_rate": 1.7686791258743473e-05, + "loss": 2.2842, "step": 2128 }, { - "epoch": 0.6818254603682946, - "grad_norm": 0.41796875, - "learning_rate": 1.3896310322747957e-05, - "loss": 2.5186, + "epoch": 0.827177581894563, + "grad_norm": 0.59765625, + "learning_rate": 1.76098064619587e-05, + "loss": 2.2952, "step": 2129 }, { - "epoch": 0.6821457165732586, - "grad_norm": 0.41796875, - "learning_rate": 1.3871266352396656e-05, - "loss": 2.1431, + "epoch": 0.8275661105849785, + "grad_norm": 0.625, + "learning_rate": 1.7532973395786467e-05, + "loss": 2.3243, "step": 2130 }, { - "epoch": 0.6824659727782226, - "grad_norm": 0.75390625, - "learning_rate": 1.3846236302133023e-05, - "loss": 2.2754, + "epoch": 0.827954639275394, + "grad_norm": 0.63671875, + "learning_rate": 1.7456292201723325e-05, + "loss": 2.3563, "step": 2131 }, { - "epoch": 0.6827862289831865, - "grad_norm": 0.34375, - "learning_rate": 1.3821220203265223e-05, - "loss": 2.122, + "epoch": 0.8283431679658094, + "grad_norm": 0.61328125, + "learning_rate": 1.7379763020986117e-05, + "loss": 2.3241, "step": 2132 }, { - "epoch": 0.6831064851881505, - "grad_norm": 0.58984375, - "learning_rate": 1.3796218087084e-05, - "loss": 2.2143, + "epoch": 0.8287316966562249, + "grad_norm": 0.61328125, + "learning_rate": 1.7303385994511778e-05, + "loss": 2.3951, "step": 2133 }, { - "epoch": 0.6834267413931145, - "grad_norm": 0.51953125, - "learning_rate": 1.3771229984862608e-05, - "loss": 2.6696, + "epoch": 0.8291202253466404, + "grad_norm": 0.62109375, + "learning_rate": 1.722716126295699e-05, + "loss": 2.3436, "step": 2134 }, { - "epoch": 0.6837469975980784, - "grad_norm": 0.50390625, - "learning_rate": 1.374625592785677e-05, - "loss": 2.481, + "epoch": 0.8295087540370559, + "grad_norm": 0.609375, + "learning_rate": 1.7151088966698004e-05, + "loss": 2.2888, "step": 2135 }, { - "epoch": 0.6840672538030425, - "grad_norm": 0.376953125, - "learning_rate": 1.3721295947304607e-05, - "loss": 2.4655, + "epoch": 0.8298972827274714, + "grad_norm": 0.6171875, + "learning_rate": 1.7075169245830348e-05, + "loss": 2.3908, "step": 2136 }, { - "epoch": 0.6843875100080064, - "grad_norm": 0.443359375, - "learning_rate": 1.3696350074426708e-05, - "loss": 2.3388, + "epoch": 0.8302858114178869, + "grad_norm": 0.66796875, + "learning_rate": 1.6999402240168505e-05, + "loss": 2.4087, "step": 2137 }, { - "epoch": 0.6847077662129704, - "grad_norm": 0.58984375, - "learning_rate": 1.3671418340425928e-05, - "loss": 2.2813, + "epoch": 0.8306743401083023, + "grad_norm": 0.60546875, + "learning_rate": 1.6923788089245762e-05, + "loss": 2.3647, "step": 2138 }, { - "epoch": 0.6850280224179344, - "grad_norm": 0.466796875, - "learning_rate": 1.3646500776487519e-05, - "loss": 2.7145, + "epoch": 0.8310628687987178, + "grad_norm": 0.6171875, + "learning_rate": 1.684832693231393e-05, + "loss": 2.3629, "step": 2139 }, { - "epoch": 0.6853482786228983, - "grad_norm": 0.515625, - "learning_rate": 1.3621597413778939e-05, - "loss": 2.4922, + "epoch": 0.8314513974891333, + "grad_norm": 0.6171875, + "learning_rate": 1.6773018908343018e-05, + "loss": 2.3733, "step": 2140 }, { - "epoch": 0.6856685348278623, - "grad_norm": 0.86328125, - "learning_rate": 1.3596708283449921e-05, - "loss": 2.0813, + "epoch": 0.8318399261795488, + "grad_norm": 0.64453125, + "learning_rate": 1.669786415602105e-05, + "loss": 2.3067, "step": 2141 }, { - "epoch": 0.6859887910328263, - "grad_norm": 0.5078125, - "learning_rate": 1.3571833416632396e-05, - "loss": 1.8429, + "epoch": 0.8322284548699643, + "grad_norm": 0.6328125, + "learning_rate": 1.662286281375377e-05, + "loss": 2.3411, "step": 2142 }, { - "epoch": 0.6863090472377902, - "grad_norm": 1.0546875, - "learning_rate": 1.3546972844440441e-05, - "loss": 2.3635, + "epoch": 0.8326169835603798, + "grad_norm": 0.6171875, + "learning_rate": 1.654801501966442e-05, + "loss": 2.3791, "step": 2143 }, { - "epoch": 0.6866293034427542, - "grad_norm": 0.494140625, - "learning_rate": 1.3522126597970244e-05, - "loss": 2.161, + "epoch": 0.8330055122507952, + "grad_norm": 0.7890625, + "learning_rate": 1.6473320911593448e-05, + "loss": 2.3025, "step": 2144 }, { - "epoch": 0.6869495596477182, - "grad_norm": 0.546875, - "learning_rate": 1.3497294708300118e-05, - "loss": 2.2284, + "epoch": 0.8333940409412107, + "grad_norm": 0.625, + "learning_rate": 1.63987806270983e-05, + "loss": 2.3023, "step": 2145 }, { - "epoch": 0.6872698158526821, - "grad_norm": 0.51171875, - "learning_rate": 1.3472477206490355e-05, - "loss": 2.4539, + "epoch": 0.8337825696316262, + "grad_norm": 0.58984375, + "learning_rate": 1.6324394303453073e-05, + "loss": 2.3044, "step": 2146 }, { - "epoch": 0.6875900720576461, - "grad_norm": 0.5703125, - "learning_rate": 1.34476741235833e-05, - "loss": 2.5645, + "epoch": 0.8341710983220417, + "grad_norm": 0.58984375, + "learning_rate": 1.625016207764839e-05, + "loss": 2.3697, "step": 2147 }, { - "epoch": 0.6879103282626101, - "grad_norm": 0.44921875, - "learning_rate": 1.3422885490603244e-05, - "loss": 2.0052, + "epoch": 0.8345596270124572, + "grad_norm": 0.59375, + "learning_rate": 1.6176084086391074e-05, + "loss": 2.3098, "step": 2148 }, { - "epoch": 0.688230584467574, - "grad_norm": 0.37890625, - "learning_rate": 1.3398111338556412e-05, - "loss": 1.8642, + "epoch": 0.8349481557028727, + "grad_norm": 0.61328125, + "learning_rate": 1.6102160466103898e-05, + "loss": 2.3355, "step": 2149 }, { - "epoch": 0.688550840672538, - "grad_norm": 0.55859375, - "learning_rate": 1.3373351698430884e-05, - "loss": 2.543, + "epoch": 0.8353366843932881, + "grad_norm": 0.625, + "learning_rate": 1.6028391352925354e-05, + "loss": 2.3012, "step": 2150 }, { - "epoch": 0.688871096877502, - "grad_norm": 0.462890625, - "learning_rate": 1.3348606601196648e-05, - "loss": 2.08, + "epoch": 0.8357252130837036, + "grad_norm": 0.6171875, + "learning_rate": 1.5954776882709365e-05, + "loss": 2.3258, "step": 2151 }, { - "epoch": 0.6891913530824659, - "grad_norm": 0.4453125, - "learning_rate": 1.332387607780543e-05, - "loss": 1.7781, + "epoch": 0.8361137417741191, + "grad_norm": 0.609375, + "learning_rate": 1.5881317191025113e-05, + "loss": 2.339, "step": 2152 }, { - "epoch": 0.6895116092874299, - "grad_norm": 0.5859375, - "learning_rate": 1.3299160159190801e-05, - "loss": 2.3629, + "epoch": 0.8365022704645346, + "grad_norm": 0.63671875, + "learning_rate": 1.5808012413156713e-05, + "loss": 2.3705, "step": 2153 }, { - "epoch": 0.689831865492394, - "grad_norm": 0.59375, - "learning_rate": 1.3274458876267998e-05, - "loss": 2.1964, + "epoch": 0.8368907991549501, + "grad_norm": 0.609375, + "learning_rate": 1.5734862684102934e-05, + "loss": 2.2996, "step": 2154 }, { - "epoch": 0.6901521216973578, - "grad_norm": 0.416015625, - "learning_rate": 1.324977225993399e-05, - "loss": 1.7751, + "epoch": 0.8372793278453656, + "grad_norm": 0.58984375, + "learning_rate": 1.5661868138577096e-05, + "loss": 2.3147, "step": 2155 }, { - "epoch": 0.6904723779023219, - "grad_norm": 0.51953125, - "learning_rate": 1.3225100341067393e-05, - "loss": 1.352, + "epoch": 0.837667856535781, + "grad_norm": 0.58984375, + "learning_rate": 1.5589028911006675e-05, + "loss": 2.3608, "step": 2156 }, { - "epoch": 0.6907926341072859, - "grad_norm": 0.56640625, - "learning_rate": 1.3200443150528438e-05, - "loss": 1.7647, + "epoch": 0.8380563852261965, + "grad_norm": 0.62109375, + "learning_rate": 1.5516345135533138e-05, + "loss": 2.4363, "step": 2157 }, { - "epoch": 0.6911128903122498, - "grad_norm": 0.88671875, - "learning_rate": 1.317580071915891e-05, - "loss": 1.7487, + "epoch": 0.838444913916612, + "grad_norm": 0.6171875, + "learning_rate": 1.5443816946011657e-05, + "loss": 2.35, "step": 2158 }, { - "epoch": 0.6914331465172138, - "grad_norm": 0.83203125, - "learning_rate": 1.3151173077782192e-05, - "loss": 2.8391, + "epoch": 0.8388334426070275, + "grad_norm": 0.609375, + "learning_rate": 1.53714444760109e-05, + "loss": 2.4453, "step": 2159 }, { - "epoch": 0.6917534027221778, - "grad_norm": 0.60546875, - "learning_rate": 1.3126560257203097e-05, - "loss": 2.3178, + "epoch": 0.839221971297443, + "grad_norm": 0.59375, + "learning_rate": 1.5299227858812693e-05, + "loss": 2.4134, "step": 2160 }, { - "epoch": 0.6920736589271417, - "grad_norm": 0.39453125, - "learning_rate": 1.3101962288207948e-05, - "loss": 1.708, + "epoch": 0.8396104999878585, + "grad_norm": 0.62109375, + "learning_rate": 1.5227167227411909e-05, + "loss": 2.3808, "step": 2161 }, { - "epoch": 0.6923939151321057, - "grad_norm": 1.078125, - "learning_rate": 1.3077379201564472e-05, - "loss": 1.5599, + "epoch": 0.8399990286782739, + "grad_norm": 0.6171875, + "learning_rate": 1.5155262714516138e-05, + "loss": 2.446, "step": 2162 }, { - "epoch": 0.6927141713370697, - "grad_norm": 0.4140625, - "learning_rate": 1.3052811028021788e-05, - "loss": 2.1572, + "epoch": 0.8403875573686894, + "grad_norm": 0.63671875, + "learning_rate": 1.508351445254549e-05, + "loss": 2.4715, "step": 2163 }, { - "epoch": 0.6930344275420336, - "grad_norm": 0.59765625, - "learning_rate": 1.3028257798310356e-05, - "loss": 2.1006, + "epoch": 0.8407760860591049, + "grad_norm": 0.640625, + "learning_rate": 1.501192257363222e-05, + "loss": 2.3122, "step": 2164 }, { - "epoch": 0.6933546837469976, - "grad_norm": 0.341796875, - "learning_rate": 1.3003719543141957e-05, - "loss": 2.0096, + "epoch": 0.8411646147495204, + "grad_norm": 0.640625, + "learning_rate": 1.494048720962069e-05, + "loss": 2.3619, "step": 2165 }, { - "epoch": 0.6936749399519616, - "grad_norm": 3.328125, - "learning_rate": 1.29791962932096e-05, - "loss": 2.0032, + "epoch": 0.8415531434399359, + "grad_norm": 0.65234375, + "learning_rate": 1.4869208492066989e-05, + "loss": 2.363, "step": 2166 }, { - "epoch": 0.6939951961569255, - "grad_norm": 0.46875, - "learning_rate": 1.29546880791876e-05, - "loss": 2.5821, + "epoch": 0.8419416721303514, + "grad_norm": 0.61328125, + "learning_rate": 1.4798086552238732e-05, + "loss": 2.3153, "step": 2167 }, { - "epoch": 0.6943154523618895, - "grad_norm": 1.1796875, - "learning_rate": 1.2930194931731382e-05, - "loss": 2.3261, + "epoch": 0.8423302008207668, + "grad_norm": 0.609375, + "learning_rate": 1.4727121521114784e-05, + "loss": 2.2836, "step": 2168 }, { - "epoch": 0.6946357085668535, - "grad_norm": 0.462890625, - "learning_rate": 1.290571688147758e-05, - "loss": 2.4759, + "epoch": 0.8427187295111823, + "grad_norm": 0.625, + "learning_rate": 1.4656313529385068e-05, + "loss": 2.345, "step": 2169 }, { - "epoch": 0.6949559647718174, - "grad_norm": 0.84765625, - "learning_rate": 1.2881253959043926e-05, - "loss": 2.61, + "epoch": 0.8431072582015978, + "grad_norm": 0.62890625, + "learning_rate": 1.4585662707450299e-05, + "loss": 2.3396, "step": 2170 }, { - "epoch": 0.6952762209767814, - "grad_norm": 0.5234375, - "learning_rate": 1.2856806195029241e-05, - "loss": 2.443, + "epoch": 0.8434957868920133, + "grad_norm": 0.61328125, + "learning_rate": 1.4515169185421751e-05, + "loss": 2.4191, "step": 2171 }, { - "epoch": 0.6955964771817454, - "grad_norm": 0.44140625, - "learning_rate": 1.2832373620013352e-05, - "loss": 2.3406, + "epoch": 0.8438843155824288, + "grad_norm": 0.609375, + "learning_rate": 1.4444833093121025e-05, + "loss": 2.3053, "step": 2172 }, { - "epoch": 0.6959167333867093, - "grad_norm": 0.54296875, - "learning_rate": 1.280795626455714e-05, - "loss": 2.389, + "epoch": 0.8442728442728443, + "grad_norm": 0.61328125, + "learning_rate": 1.4374654560079725e-05, + "loss": 2.2813, "step": 2173 }, { - "epoch": 0.6962369895916733, - "grad_norm": 0.4140625, - "learning_rate": 1.2783554159202404e-05, - "loss": 2.4067, + "epoch": 0.8446613729632597, + "grad_norm": 0.59765625, + "learning_rate": 1.4304633715539384e-05, + "loss": 2.3522, "step": 2174 }, { - "epoch": 0.6965572457966374, - "grad_norm": 0.443359375, - "learning_rate": 1.2759167334471884e-05, - "loss": 2.178, + "epoch": 0.8450499016536752, + "grad_norm": 0.62109375, + "learning_rate": 1.4234770688451082e-05, + "loss": 2.4164, "step": 2175 }, { - "epoch": 0.6968775020016013, - "grad_norm": 0.373046875, - "learning_rate": 1.2734795820869208e-05, - "loss": 2.0938, + "epoch": 0.8454384303440907, + "grad_norm": 0.61328125, + "learning_rate": 1.416506560747528e-05, + "loss": 2.3329, "step": 2176 }, { - "epoch": 0.6971977582065653, - "grad_norm": 0.4296875, - "learning_rate": 1.2710439648878855e-05, - "loss": 1.9665, + "epoch": 0.8458269590345062, + "grad_norm": 0.61328125, + "learning_rate": 1.4095518600981562e-05, + "loss": 2.3285, "step": 2177 }, { - "epoch": 0.6975180144115293, - "grad_norm": 0.5390625, - "learning_rate": 1.2686098848966107e-05, - "loss": 2.3108, + "epoch": 0.8462154877249217, + "grad_norm": 0.62109375, + "learning_rate": 1.4026129797048393e-05, + "loss": 2.2963, "step": 2178 }, { - "epoch": 0.6978382706164932, - "grad_norm": 0.4453125, - "learning_rate": 1.2661773451577036e-05, - "loss": 1.9203, + "epoch": 0.8466040164153372, + "grad_norm": 0.6640625, + "learning_rate": 1.3956899323462891e-05, + "loss": 2.3549, "step": 2179 }, { - "epoch": 0.6981585268214572, - "grad_norm": 0.91015625, - "learning_rate": 1.2637463487138398e-05, - "loss": 2.6032, + "epoch": 0.8469925451057526, + "grad_norm": 0.609375, + "learning_rate": 1.3887827307720636e-05, + "loss": 2.3483, "step": 2180 }, { - "epoch": 0.6984787830264212, - "grad_norm": 0.65625, - "learning_rate": 1.2613168986057728e-05, - "loss": 2.3923, + "epoch": 0.8473810737961681, + "grad_norm": 0.62109375, + "learning_rate": 1.3818913877025286e-05, + "loss": 2.4284, "step": 2181 }, { - "epoch": 0.6987990392313851, - "grad_norm": 0.71484375, - "learning_rate": 1.2588889978723134e-05, - "loss": 1.8835, + "epoch": 0.8477696024865836, + "grad_norm": 0.60546875, + "learning_rate": 1.3750159158288546e-05, + "loss": 2.3375, "step": 2182 }, { - "epoch": 0.6991192954363491, - "grad_norm": 0.42578125, - "learning_rate": 1.2564626495503401e-05, - "loss": 1.7873, + "epoch": 0.8481581311769991, + "grad_norm": 0.58984375, + "learning_rate": 1.3681563278129794e-05, + "loss": 2.3789, "step": 2183 }, { - "epoch": 0.699439551641313, - "grad_norm": 0.4921875, - "learning_rate": 1.2540378566747874e-05, - "loss": 1.8807, + "epoch": 0.8485466598674146, + "grad_norm": 0.609375, + "learning_rate": 1.3613126362875906e-05, + "loss": 2.3979, "step": 2184 }, { - "epoch": 0.699759807846277, - "grad_norm": 0.5625, - "learning_rate": 1.2516146222786451e-05, - "loss": 2.3201, + "epoch": 0.8489351885578301, + "grad_norm": 0.62890625, + "learning_rate": 1.3544848538560972e-05, + "loss": 2.3132, "step": 2185 }, { - "epoch": 0.700080064051241, - "grad_norm": 0.396484375, - "learning_rate": 1.2491929493929508e-05, - "loss": 1.9649, + "epoch": 0.8493237172482455, + "grad_norm": 0.6484375, + "learning_rate": 1.3476729930926147e-05, + "loss": 2.3686, "step": 2186 }, { - "epoch": 0.7004003202562049, - "grad_norm": 0.55078125, - "learning_rate": 1.2467728410467944e-05, - "loss": 1.8792, + "epoch": 0.849712245938661, + "grad_norm": 0.6171875, + "learning_rate": 1.3408770665419335e-05, + "loss": 2.4061, "step": 2187 }, { - "epoch": 0.7007205764611689, + "epoch": 0.8501007746290765, "grad_norm": 0.6171875, - "learning_rate": 1.2443543002673022e-05, - "loss": 2.3325, + "learning_rate": 1.3340970867195013e-05, + "loss": 2.3404, "step": 2188 }, { - "epoch": 0.7010408326661329, - "grad_norm": 0.73828125, - "learning_rate": 1.241937330079647e-05, - "loss": 2.6154, + "epoch": 0.850489303319492, + "grad_norm": 0.58984375, + "learning_rate": 1.3273330661113992e-05, + "loss": 2.3503, "step": 2189 }, { - "epoch": 0.7013610888710968, - "grad_norm": 0.44921875, - "learning_rate": 1.2395219335070301e-05, - "loss": 2.442, + "epoch": 0.8508778320099075, + "grad_norm": 0.625, + "learning_rate": 1.3205850171743106e-05, + "loss": 2.3556, "step": 2190 }, { - "epoch": 0.7016813450760608, - "grad_norm": 0.46484375, - "learning_rate": 1.2371081135706888e-05, - "loss": 2.3294, + "epoch": 0.8512663607003229, + "grad_norm": 0.61328125, + "learning_rate": 1.3138529523355148e-05, + "loss": 2.3403, "step": 2191 }, { - "epoch": 0.7020016012810248, - "grad_norm": 0.5625, - "learning_rate": 1.234695873289887e-05, - "loss": 2.4601, + "epoch": 0.8516548893907384, + "grad_norm": 0.6015625, + "learning_rate": 1.3071368839928488e-05, + "loss": 2.3289, "step": 2192 }, { - "epoch": 0.7023218574859887, - "grad_norm": 0.490234375, - "learning_rate": 1.2322852156819137e-05, - "loss": 2.0798, + "epoch": 0.8520434180811539, + "grad_norm": 0.6171875, + "learning_rate": 1.3004368245146915e-05, + "loss": 2.3767, "step": 2193 }, { - "epoch": 0.7026421136909528, - "grad_norm": 0.4453125, - "learning_rate": 1.2298761437620748e-05, - "loss": 2.283, + "epoch": 0.8524319467715694, + "grad_norm": 0.62890625, + "learning_rate": 1.2937527862399424e-05, + "loss": 2.3762, "step": 2194 }, { - "epoch": 0.7029623698959168, - "grad_norm": 0.498046875, - "learning_rate": 1.2274686605436989e-05, - "loss": 2.3894, + "epoch": 0.8528204754619849, + "grad_norm": 0.59375, + "learning_rate": 1.2870847814779907e-05, + "loss": 2.3307, "step": 2195 }, { - "epoch": 0.7032826261008807, - "grad_norm": 0.53125, - "learning_rate": 1.2250627690381213e-05, - "loss": 2.4578, + "epoch": 0.8532090041524004, + "grad_norm": 0.609375, + "learning_rate": 1.2804328225087048e-05, + "loss": 2.3917, "step": 2196 }, { - "epoch": 0.7036028823058447, - "grad_norm": 0.52734375, - "learning_rate": 1.2226584722546899e-05, - "loss": 2.2, + "epoch": 0.8535975328428158, + "grad_norm": 0.6328125, + "learning_rate": 1.2737969215823986e-05, + "loss": 2.4013, "step": 2197 }, { - "epoch": 0.7039231385108087, - "grad_norm": 0.412109375, - "learning_rate": 1.2202557732007569e-05, - "loss": 1.6341, + "epoch": 0.8539860615332313, + "grad_norm": 0.6015625, + "learning_rate": 1.2671770909198122e-05, + "loss": 2.3378, "step": 2198 }, { - "epoch": 0.7042433947157726, - "grad_norm": 0.58984375, - "learning_rate": 1.217854674881677e-05, - "loss": 2.0122, + "epoch": 0.8543745902236468, + "grad_norm": 0.61328125, + "learning_rate": 1.2605733427120925e-05, + "loss": 2.3566, "step": 2199 }, { - "epoch": 0.7045636509207366, - "grad_norm": 0.5703125, - "learning_rate": 1.2154551803007986e-05, - "loss": 2.0305, + "epoch": 0.8547631189140623, + "grad_norm": 0.59375, + "learning_rate": 1.2539856891207712e-05, + "loss": 2.4097, "step": 2200 }, { - "epoch": 0.7048839071257006, - "grad_norm": 0.93359375, - "learning_rate": 1.2130572924594713e-05, - "loss": 1.7452, + "epoch": 0.8551516476044778, + "grad_norm": 0.6171875, + "learning_rate": 1.2474141422777363e-05, + "loss": 2.3139, "step": 2201 }, { - "epoch": 0.7052041633306645, - "grad_norm": 0.5625, - "learning_rate": 1.2106610143570271e-05, - "loss": 1.9185, + "epoch": 0.8555401762948933, + "grad_norm": 0.625, + "learning_rate": 1.2408587142852179e-05, + "loss": 2.3955, "step": 2202 }, { - "epoch": 0.7055244195356285, - "grad_norm": 0.494140625, - "learning_rate": 1.208266348990792e-05, - "loss": 2.1781, + "epoch": 0.8559287049853087, + "grad_norm": 0.5859375, + "learning_rate": 1.2343194172157535e-05, + "loss": 2.3451, "step": 2203 }, { - "epoch": 0.7058446757405925, - "grad_norm": 0.59765625, - "learning_rate": 1.2058732993560678e-05, - "loss": 2.6887, + "epoch": 0.8563172336757242, + "grad_norm": 0.60546875, + "learning_rate": 1.2277962631121809e-05, + "loss": 2.3197, "step": 2204 }, { - "epoch": 0.7061649319455564, - "grad_norm": 0.53515625, - "learning_rate": 1.2034818684461402e-05, - "loss": 2.3117, + "epoch": 0.8567057623661397, + "grad_norm": 0.58203125, + "learning_rate": 1.221289263987606e-05, + "loss": 2.319, "step": 2205 }, { - "epoch": 0.7064851881505204, - "grad_norm": 0.46484375, - "learning_rate": 1.2010920592522681e-05, - "loss": 2.0872, + "epoch": 0.8570942910565552, + "grad_norm": 0.59375, + "learning_rate": 1.2147984318253857e-05, + "loss": 2.3246, "step": 2206 }, { - "epoch": 0.7068054443554844, - "grad_norm": 1.296875, - "learning_rate": 1.198703874763683e-05, - "loss": 2.1383, + "epoch": 0.8574828197469707, + "grad_norm": 0.59375, + "learning_rate": 1.2083237785791003e-05, + "loss": 2.3354, "step": 2207 }, { - "epoch": 0.7071257005604483, - "grad_norm": 1.625, - "learning_rate": 1.1963173179675812e-05, - "loss": 2.1324, + "epoch": 0.8578713484373862, + "grad_norm": 0.65234375, + "learning_rate": 1.201865316172539e-05, + "loss": 2.4292, "step": 2208 }, { - "epoch": 0.7074459567654123, - "grad_norm": 0.515625, - "learning_rate": 1.1939323918491288e-05, - "loss": 2.2598, + "epoch": 0.8582598771278016, + "grad_norm": 0.60546875, + "learning_rate": 1.1954230564996682e-05, + "loss": 2.3665, "step": 2209 }, { - "epoch": 0.7077662129703763, - "grad_norm": 0.84765625, - "learning_rate": 1.1915490993914458e-05, - "loss": 2.0589, + "epoch": 0.8586484058182171, + "grad_norm": 0.60546875, + "learning_rate": 1.1889970114246196e-05, + "loss": 2.2845, "step": 2210 }, { - "epoch": 0.7080864691753402, - "grad_norm": 0.7109375, - "learning_rate": 1.1891674435756129e-05, - "loss": 2.522, + "epoch": 0.8590369345086326, + "grad_norm": 0.6171875, + "learning_rate": 1.1825871927816635e-05, + "loss": 2.3748, "step": 2211 }, { - "epoch": 0.7084067253803042, - "grad_norm": 0.53515625, - "learning_rate": 1.1867874273806623e-05, - "loss": 2.4149, + "epoch": 0.8594254631990481, + "grad_norm": 0.6171875, + "learning_rate": 1.176193612375186e-05, + "loss": 2.3349, "step": 2212 }, { - "epoch": 0.7087269815852683, - "grad_norm": 1.140625, - "learning_rate": 1.1844090537835764e-05, - "loss": 2.5084, + "epoch": 0.8598139918894636, + "grad_norm": 0.625, + "learning_rate": 1.1698162819796698e-05, + "loss": 2.3285, "step": 2213 }, { - "epoch": 0.7090472377902322, - "grad_norm": 0.6328125, - "learning_rate": 1.1820323257592796e-05, - "loss": 2.5254, + "epoch": 0.8602025205798791, + "grad_norm": 0.59375, + "learning_rate": 1.1634552133396704e-05, + "loss": 2.3757, "step": 2214 }, { - "epoch": 0.7093674939951962, - "grad_norm": 0.73828125, - "learning_rate": 1.1796572462806437e-05, - "loss": 2.2132, + "epoch": 0.8605910492702945, + "grad_norm": 0.5859375, + "learning_rate": 1.1571104181697957e-05, + "loss": 2.3622, "step": 2215 }, { - "epoch": 0.7096877502001602, - "grad_norm": 0.5078125, - "learning_rate": 1.177283818318472e-05, - "loss": 2.0243, + "epoch": 0.86097957796071, + "grad_norm": 0.625, + "learning_rate": 1.1507819081546878e-05, + "loss": 2.3709, "step": 2216 }, { - "epoch": 0.7100080064051241, - "grad_norm": 0.640625, - "learning_rate": 1.1749120448415085e-05, - "loss": 2.1482, + "epoch": 0.8613681066511255, + "grad_norm": 0.59375, + "learning_rate": 1.1444696949489908e-05, + "loss": 2.3431, "step": 2217 }, { - "epoch": 0.7103282626100881, - "grad_norm": 0.466796875, - "learning_rate": 1.1725419288164222e-05, - "loss": 1.9561, + "epoch": 0.861756635341541, + "grad_norm": 0.59765625, + "learning_rate": 1.1381737901773405e-05, + "loss": 2.3009, "step": 2218 }, { - "epoch": 0.7106485188150521, - "grad_norm": 0.5078125, - "learning_rate": 1.1701734732078118e-05, - "loss": 2.4484, + "epoch": 0.8621451640319565, + "grad_norm": 0.61328125, + "learning_rate": 1.1318942054343395e-05, + "loss": 2.3848, "step": 2219 }, { - "epoch": 0.710968775020016, - "grad_norm": 0.7109375, - "learning_rate": 1.1678066809781988e-05, - "loss": 2.4673, + "epoch": 0.862533692722372, + "grad_norm": 0.62109375, + "learning_rate": 1.1256309522845355e-05, + "loss": 2.3062, "step": 2220 }, { - "epoch": 0.71128903122498, - "grad_norm": 0.3515625, - "learning_rate": 1.1654415550880243e-05, - "loss": 2.1008, + "epoch": 0.8629222214127874, + "grad_norm": 0.6484375, + "learning_rate": 1.1193840422623981e-05, + "loss": 2.3867, "step": 2221 }, { - "epoch": 0.711609287429944, - "grad_norm": 0.6875, - "learning_rate": 1.163078098495642e-05, - "loss": 2.556, + "epoch": 0.8633107501032029, + "grad_norm": 0.6328125, + "learning_rate": 1.1131534868722993e-05, + "loss": 2.4272, "step": 2222 }, { - "epoch": 0.7119295436349079, - "grad_norm": 0.478515625, - "learning_rate": 1.1607163141573246e-05, - "loss": 2.6396, + "epoch": 0.8636992787936184, + "grad_norm": 0.60546875, + "learning_rate": 1.106939297588494e-05, + "loss": 2.3541, "step": 2223 }, { - "epoch": 0.7122497998398719, - "grad_norm": 0.55859375, - "learning_rate": 1.158356205027245e-05, - "loss": 2.5558, + "epoch": 0.8640878074840339, + "grad_norm": 0.6015625, + "learning_rate": 1.1007414858550968e-05, + "loss": 2.3333, "step": 2224 }, { - "epoch": 0.7125700560448359, - "grad_norm": 0.45703125, - "learning_rate": 1.1559977740574861e-05, - "loss": 2.4041, + "epoch": 0.8644763361744494, + "grad_norm": 0.62109375, + "learning_rate": 1.0945600630860563e-05, + "loss": 2.3475, "step": 2225 }, { - "epoch": 0.7128903122497998, - "grad_norm": 0.609375, - "learning_rate": 1.1536410241980297e-05, - "loss": 2.3616, + "epoch": 0.8648648648648649, + "grad_norm": 0.6171875, + "learning_rate": 1.0883950406651466e-05, + "loss": 2.354, "step": 2226 }, { - "epoch": 0.7132105684547638, - "grad_norm": 0.56640625, - "learning_rate": 1.1512859583967552e-05, - "loss": 2.4075, + "epoch": 0.8652533935552803, + "grad_norm": 0.65625, + "learning_rate": 1.0822464299459345e-05, + "loss": 2.3314, "step": 2227 }, { - "epoch": 0.7135308246597278, - "grad_norm": 0.431640625, - "learning_rate": 1.148932579599436e-05, - "loss": 2.4303, + "epoch": 0.8656419222456958, + "grad_norm": 0.58984375, + "learning_rate": 1.0761142422517623e-05, + "loss": 2.2504, "step": 2228 }, { - "epoch": 0.7138510808646917, - "grad_norm": 0.6484375, - "learning_rate": 1.1465808907497352e-05, - "loss": 2.5852, + "epoch": 0.8660304509361113, + "grad_norm": 0.59765625, + "learning_rate": 1.0699984888757308e-05, + "loss": 2.3928, "step": 2229 }, { - "epoch": 0.7141713370696557, - "grad_norm": 1.5859375, - "learning_rate": 1.1442308947891992e-05, - "loss": 2.5451, + "epoch": 0.8664189796265268, + "grad_norm": 0.62109375, + "learning_rate": 1.063899181080672e-05, + "loss": 2.345, "step": 2230 }, { - "epoch": 0.7144915932746196, - "grad_norm": 0.61328125, - "learning_rate": 1.141882594657263e-05, - "loss": 1.6112, + "epoch": 0.8668075083169423, + "grad_norm": 0.6015625, + "learning_rate": 1.0578163300991328e-05, + "loss": 2.3404, "step": 2231 }, { - "epoch": 0.7148118494795836, - "grad_norm": 0.5234375, - "learning_rate": 1.1395359932912348e-05, - "loss": 2.3431, + "epoch": 0.8671960370073578, + "grad_norm": 0.609375, + "learning_rate": 1.0517499471333536e-05, + "loss": 2.3495, "step": 2232 }, { - "epoch": 0.7151321056845477, - "grad_norm": 8.1875, - "learning_rate": 1.1371910936263e-05, - "loss": 2.2054, + "epoch": 0.8675845656977732, + "grad_norm": 0.61328125, + "learning_rate": 1.0457000433552477e-05, + "loss": 2.3391, "step": 2233 }, { - "epoch": 0.7154523618895116, - "grad_norm": 0.365234375, - "learning_rate": 1.1348478985955161e-05, - "loss": 1.7845, + "epoch": 0.8679730943881887, + "grad_norm": 0.59375, + "learning_rate": 1.0396666299063763e-05, + "loss": 2.362, "step": 2234 }, { - "epoch": 0.7157726180944756, - "grad_norm": 0.46875, - "learning_rate": 1.1325064111298087e-05, - "loss": 2.5106, + "epoch": 0.8683616230786042, + "grad_norm": 0.60546875, + "learning_rate": 1.0336497178979343e-05, + "loss": 2.2885, "step": 2235 }, { - "epoch": 0.7160928742994396, - "grad_norm": 0.376953125, - "learning_rate": 1.1301666341579633e-05, - "loss": 2.3032, + "epoch": 0.8687501517690197, + "grad_norm": 0.65234375, + "learning_rate": 1.0276493184107305e-05, + "loss": 2.3794, "step": 2236 }, { - "epoch": 0.7164131305044035, - "grad_norm": 0.498046875, - "learning_rate": 1.1278285706066335e-05, - "loss": 2.0377, + "epoch": 0.8691386804594352, + "grad_norm": 0.6171875, + "learning_rate": 1.0216654424951589e-05, + "loss": 2.3564, "step": 2237 }, { - "epoch": 0.7167333867093675, - "grad_norm": 0.51953125, - "learning_rate": 1.1254922234003224e-05, - "loss": 2.2911, + "epoch": 0.8695272091498507, + "grad_norm": 0.61328125, + "learning_rate": 1.0156981011711875e-05, + "loss": 2.3339, "step": 2238 }, { - "epoch": 0.7170536429143315, - "grad_norm": 0.5, - "learning_rate": 1.1231575954613902e-05, - "loss": 1.9858, + "epoch": 0.8699157378402661, + "grad_norm": 0.7265625, + "learning_rate": 1.0097473054283301e-05, + "loss": 2.3127, "step": 2239 }, { - "epoch": 0.7173738991192954, - "grad_norm": 0.7578125, - "learning_rate": 1.1208246897100458e-05, - "loss": 2.2628, + "epoch": 0.8703042665306816, + "grad_norm": 0.58984375, + "learning_rate": 1.003813066225635e-05, + "loss": 2.3678, "step": 2240 }, { - "epoch": 0.7176941553242594, - "grad_norm": 0.474609375, - "learning_rate": 1.1184935090643433e-05, - "loss": 1.7735, + "epoch": 0.8706927952210971, + "grad_norm": 0.60546875, + "learning_rate": 9.97895394491657e-06, + "loss": 2.3859, "step": 2241 }, { - "epoch": 0.7180144115292234, - "grad_norm": 0.322265625, - "learning_rate": 1.1161640564401805e-05, - "loss": 1.6539, + "epoch": 0.8710813239115126, + "grad_norm": 0.6015625, + "learning_rate": 9.91994301124437e-06, + "loss": 2.3713, "step": 2242 }, { - "epoch": 0.7183346677341873, - "grad_norm": 1.078125, - "learning_rate": 1.1138363347512926e-05, - "loss": 2.1595, + "epoch": 0.8714698526019281, + "grad_norm": 0.6171875, + "learning_rate": 9.861097969914902e-06, + "loss": 2.3683, "step": 2243 }, { - "epoch": 0.7186549239391513, - "grad_norm": 0.310546875, - "learning_rate": 1.1115103469092475e-05, - "loss": 1.9965, + "epoch": 0.8718583812923436, + "grad_norm": 0.58203125, + "learning_rate": 9.802418929297773e-06, + "loss": 2.2977, "step": 2244 }, { - "epoch": 0.7189751801441153, - "grad_norm": 0.68359375, - "learning_rate": 1.1091860958234508e-05, - "loss": 2.3327, + "epoch": 0.872246909982759, + "grad_norm": 0.5859375, + "learning_rate": 9.743905997456926e-06, + "loss": 2.3027, "step": 2245 }, { - "epoch": 0.7192954363490792, - "grad_norm": 0.4375, - "learning_rate": 1.1068635844011286e-05, - "loss": 2.1586, + "epoch": 0.8726354386731745, + "grad_norm": 0.5859375, + "learning_rate": 9.68555928215037e-06, + "loss": 2.3658, "step": 2246 }, { - "epoch": 0.7196156925540432, - "grad_norm": 0.5859375, - "learning_rate": 1.1045428155473347e-05, - "loss": 1.9614, + "epoch": 0.87302396736359, + "grad_norm": 0.61328125, + "learning_rate": 9.627378890829963e-06, + "loss": 2.2962, "step": 2247 }, { - "epoch": 0.7199359487590072, - "grad_norm": 0.4375, - "learning_rate": 1.1022237921649429e-05, - "loss": 1.8978, + "epoch": 0.8734124960540055, + "grad_norm": 0.58203125, + "learning_rate": 9.569364930641323e-06, + "loss": 2.3248, "step": 2248 }, { - "epoch": 0.7202562049639711, - "grad_norm": 0.52734375, - "learning_rate": 1.0999065171546432e-05, - "loss": 2.1097, + "epoch": 0.873801024744421, + "grad_norm": 0.59375, + "learning_rate": 9.51151750842354e-06, + "loss": 2.3288, "step": 2249 }, { - "epoch": 0.7205764611689351, - "grad_norm": 0.46875, - "learning_rate": 1.0975909934149392e-05, - "loss": 2.1752, + "epoch": 0.8741895534348365, + "grad_norm": 0.5859375, + "learning_rate": 9.453836730709031e-06, + "loss": 2.3732, "step": 2250 }, { - "epoch": 0.7208967173738992, - "grad_norm": 0.65234375, - "learning_rate": 1.0952772238421444e-05, - "loss": 2.2443, + "epoch": 0.8745780821252519, + "grad_norm": 0.62109375, + "learning_rate": 9.396322703723282e-06, + "loss": 2.3707, "step": 2251 }, { - "epoch": 0.721216973578863, - "grad_norm": 0.578125, - "learning_rate": 1.0929652113303754e-05, - "loss": 2.4343, + "epoch": 0.8749666108156674, + "grad_norm": 0.59765625, + "learning_rate": 9.338975533384719e-06, + "loss": 2.3222, "step": 2252 }, { - "epoch": 0.7215372297838271, - "grad_norm": 0.47265625, - "learning_rate": 1.0906549587715546e-05, - "loss": 2.1622, + "epoch": 0.8753551395060829, + "grad_norm": 0.60546875, + "learning_rate": 9.281795325304455e-06, + "loss": 2.3055, "step": 2253 }, { - "epoch": 0.7218574859887911, - "grad_norm": 0.515625, - "learning_rate": 1.0883464690554005e-05, - "loss": 2.4342, + "epoch": 0.8757436681964984, + "grad_norm": 0.6484375, + "learning_rate": 9.224782184786129e-06, + "loss": 2.4232, "step": 2254 }, { - "epoch": 0.722177742193755, - "grad_norm": 0.498046875, - "learning_rate": 1.0860397450694285e-05, - "loss": 2.6058, + "epoch": 0.8761321968869139, + "grad_norm": 0.6328125, + "learning_rate": 9.167936216825746e-06, + "loss": 2.3638, "step": 2255 }, { - "epoch": 0.722497998398719, - "grad_norm": 0.5703125, - "learning_rate": 1.0837347896989437e-05, - "loss": 2.2438, + "epoch": 0.8765207255773293, + "grad_norm": 0.59375, + "learning_rate": 9.111257526111394e-06, + "loss": 2.4049, "step": 2256 }, { - "epoch": 0.722818254603683, - "grad_norm": 0.58984375, - "learning_rate": 1.081431605827041e-05, - "loss": 2.1936, + "epoch": 0.8769092542677448, + "grad_norm": 0.60546875, + "learning_rate": 9.054746217023125e-06, + "loss": 2.3684, "step": 2257 }, { - "epoch": 0.7231385108086469, - "grad_norm": 0.52734375, - "learning_rate": 1.0791301963345954e-05, - "loss": 2.4717, + "epoch": 0.8772977829581603, + "grad_norm": 0.62109375, + "learning_rate": 8.998402393632754e-06, + "loss": 2.3333, "step": 2258 }, { - "epoch": 0.7234587670136109, - "grad_norm": 1.0078125, - "learning_rate": 1.0768305641002688e-05, - "loss": 2.1485, + "epoch": 0.8776863116485758, + "grad_norm": 0.609375, + "learning_rate": 8.942226159703603e-06, + "loss": 2.3723, "step": 2259 }, { - "epoch": 0.7237790232185749, - "grad_norm": 0.62109375, - "learning_rate": 1.0745327120004935e-05, - "loss": 2.4316, + "epoch": 0.8780748403389913, + "grad_norm": 0.60546875, + "learning_rate": 8.88621761869044e-06, + "loss": 2.3631, "step": 2260 }, { - "epoch": 0.7240992794235388, - "grad_norm": 0.58984375, - "learning_rate": 1.0722366429094798e-05, - "loss": 1.7809, + "epoch": 0.8784633690294068, + "grad_norm": 0.61328125, + "learning_rate": 8.83037687373911e-06, + "loss": 2.3523, "step": 2261 }, { - "epoch": 0.7244195356285028, - "grad_norm": 0.6484375, - "learning_rate": 1.0699423596992056e-05, - "loss": 2.3482, + "epoch": 0.8788518977198222, + "grad_norm": 0.62109375, + "learning_rate": 8.774704027686509e-06, + "loss": 2.3548, "step": 2262 }, { - "epoch": 0.7247397918334668, - "grad_norm": 0.453125, - "learning_rate": 1.0676498652394157e-05, - "loss": 2.037, + "epoch": 0.8792404264102377, + "grad_norm": 0.609375, + "learning_rate": 8.719199183060322e-06, + "loss": 2.3561, "step": 2263 }, { - "epoch": 0.7250600480384307, - "grad_norm": 0.419921875, - "learning_rate": 1.065359162397617e-05, - "loss": 2.3973, + "epoch": 0.8796289551006532, + "grad_norm": 0.6015625, + "learning_rate": 8.663862442078819e-06, + "loss": 2.3768, "step": 2264 }, { - "epoch": 0.7253803042433947, - "grad_norm": 0.404296875, - "learning_rate": 1.0630702540390766e-05, - "loss": 2.2461, + "epoch": 0.8800174837910687, + "grad_norm": 0.62109375, + "learning_rate": 8.608693906650711e-06, + "loss": 2.4043, "step": 2265 }, { - "epoch": 0.7257005604483587, - "grad_norm": 0.87109375, - "learning_rate": 1.060783143026814e-05, - "loss": 2.4446, + "epoch": 0.8804060124814842, + "grad_norm": 0.6171875, + "learning_rate": 8.55369367837493e-06, + "loss": 2.3483, "step": 2266 }, { - "epoch": 0.7260208166533226, - "grad_norm": 0.57421875, - "learning_rate": 1.0584978322216057e-05, - "loss": 1.8983, + "epoch": 0.8807945411718997, + "grad_norm": 0.63671875, + "learning_rate": 8.498861858540451e-06, + "loss": 2.3346, "step": 2267 }, { - "epoch": 0.7263410728582866, - "grad_norm": 0.72265625, - "learning_rate": 1.0562143244819707e-05, - "loss": 1.9454, + "epoch": 0.8811830698623151, + "grad_norm": 0.6171875, + "learning_rate": 8.444198548126104e-06, + "loss": 2.3474, "step": 2268 }, { - "epoch": 0.7266613290632506, - "grad_norm": 0.478515625, - "learning_rate": 1.0539326226641768e-05, - "loss": 2.2344, + "epoch": 0.8815715985527306, + "grad_norm": 0.58203125, + "learning_rate": 8.389703847800402e-06, + "loss": 2.3003, "step": 2269 }, { - "epoch": 0.7269815852682145, - "grad_norm": 0.46484375, - "learning_rate": 1.0516527296222308e-05, - "loss": 2.0809, + "epoch": 0.8819601272431461, + "grad_norm": 0.59375, + "learning_rate": 8.335377857921322e-06, + "loss": 2.3599, "step": 2270 }, { - "epoch": 0.7273018414731786, - "grad_norm": 0.4765625, - "learning_rate": 1.0493746482078783e-05, - "loss": 1.7608, + "epoch": 0.8823486559335616, + "grad_norm": 0.60546875, + "learning_rate": 8.28122067853615e-06, + "loss": 2.3675, "step": 2271 }, { - "epoch": 0.7276220976781426, - "grad_norm": 0.455078125, - "learning_rate": 1.0470983812705984e-05, - "loss": 2.4922, + "epoch": 0.8827371846239771, + "grad_norm": 0.61328125, + "learning_rate": 8.227232409381314e-06, + "loss": 2.3494, "step": 2272 }, { - "epoch": 0.7279423538831065, - "grad_norm": 0.33984375, - "learning_rate": 1.0448239316576009e-05, - "loss": 1.822, + "epoch": 0.8831257133143926, + "grad_norm": 0.5859375, + "learning_rate": 8.173413149882147e-06, + "loss": 2.2979, "step": 2273 }, { - "epoch": 0.7282626100880705, - "grad_norm": 0.376953125, - "learning_rate": 1.0425513022138203e-05, - "loss": 2.2417, + "epoch": 0.883514242004808, + "grad_norm": 0.61328125, + "learning_rate": 8.119762999152747e-06, + "loss": 2.3499, "step": 2274 }, { - "epoch": 0.7285828662930345, - "grad_norm": 0.5234375, - "learning_rate": 1.0402804957819174e-05, - "loss": 1.8278, + "epoch": 0.8839027706952235, + "grad_norm": 0.6015625, + "learning_rate": 8.066282055995766e-06, + "loss": 2.2933, "step": 2275 }, { - "epoch": 0.7289031224979984, - "grad_norm": 0.62890625, - "learning_rate": 1.0380115152022716e-05, - "loss": 2.0429, + "epoch": 0.884291299385639, + "grad_norm": 0.63671875, + "learning_rate": 8.012970418902244e-06, + "loss": 2.3571, "step": 2276 }, { - "epoch": 0.7292233787029624, - "grad_norm": 0.3671875, - "learning_rate": 1.0357443633129777e-05, - "loss": 2.1444, + "epoch": 0.8846798280760545, + "grad_norm": 0.59765625, + "learning_rate": 7.959828186051454e-06, + "loss": 2.3032, "step": 2277 }, { - "epoch": 0.7295436349079263, - "grad_norm": 0.53515625, - "learning_rate": 1.0334790429498445e-05, - "loss": 2.5259, + "epoch": 0.88506835676647, + "grad_norm": 0.62109375, + "learning_rate": 7.906855455310647e-06, + "loss": 2.3277, "step": 2278 }, { - "epoch": 0.7298638911128903, - "grad_norm": 0.51171875, - "learning_rate": 1.0312155569463897e-05, - "loss": 2.4972, + "epoch": 0.8854568854568855, + "grad_norm": 0.59375, + "learning_rate": 7.85405232423494e-06, + "loss": 2.3656, "step": 2279 }, { - "epoch": 0.7301841473178543, - "grad_norm": 0.5234375, - "learning_rate": 1.0289539081338335e-05, - "loss": 1.7737, + "epoch": 0.8858454141473009, + "grad_norm": 0.58203125, + "learning_rate": 7.801418890067114e-06, + "loss": 2.3602, "step": 2280 }, { - "epoch": 0.7305044035228182, - "grad_norm": 0.40234375, - "learning_rate": 1.026694099341104e-05, - "loss": 2.4937, + "epoch": 0.8862339428377164, + "grad_norm": 0.6171875, + "learning_rate": 7.748955249737432e-06, + "loss": 2.3514, "step": 2281 }, { - "epoch": 0.7308246597277822, - "grad_norm": 0.62890625, - "learning_rate": 1.024436133394822e-05, - "loss": 1.7839, + "epoch": 0.8866224715281319, + "grad_norm": 0.6171875, + "learning_rate": 7.696661499863467e-06, + "loss": 2.3756, "step": 2282 }, { - "epoch": 0.7311449159327462, - "grad_norm": 0.470703125, - "learning_rate": 1.022180013119306e-05, - "loss": 2.5823, + "epoch": 0.8870110002185474, + "grad_norm": 0.60546875, + "learning_rate": 7.644537736749924e-06, + "loss": 2.3808, "step": 2283 }, { - "epoch": 0.7314651721377101, - "grad_norm": 0.640625, - "learning_rate": 1.0199257413365663e-05, - "loss": 2.5663, + "epoch": 0.8873995289089629, + "grad_norm": 0.57421875, + "learning_rate": 7.5925840563884405e-06, + "loss": 2.3155, "step": 2284 }, { - "epoch": 0.7317854283426741, - "grad_norm": 0.49609375, - "learning_rate": 1.0176733208662993e-05, - "loss": 2.738, + "epoch": 0.8877880575993784, + "grad_norm": 0.62109375, + "learning_rate": 7.54080055445745e-06, + "loss": 2.3637, "step": 2285 }, { - "epoch": 0.7321056845476381, - "grad_norm": 0.5625, - "learning_rate": 1.0154227545258873e-05, - "loss": 2.3974, + "epoch": 0.8881765862897938, + "grad_norm": 0.6328125, + "learning_rate": 7.489187326321978e-06, + "loss": 2.2882, "step": 2286 }, { - "epoch": 0.732425940752602, - "grad_norm": 0.93359375, - "learning_rate": 1.0131740451303936e-05, - "loss": 2.6672, + "epoch": 0.8885651149802093, + "grad_norm": 0.59765625, + "learning_rate": 7.437744467033436e-06, + "loss": 2.3655, "step": 2287 }, { - "epoch": 0.732746196957566, - "grad_norm": 0.412109375, - "learning_rate": 1.0109271954925562e-05, - "loss": 1.7195, + "epoch": 0.8889536436706248, + "grad_norm": 0.59765625, + "learning_rate": 7.386472071329542e-06, + "loss": 2.4046, "step": 2288 }, { - "epoch": 0.73306645316253, - "grad_norm": 0.51171875, - "learning_rate": 1.0086822084227896e-05, - "loss": 2.2517, + "epoch": 0.8893421723610403, + "grad_norm": 0.64453125, + "learning_rate": 7.335370233634042e-06, + "loss": 2.31, "step": 2289 }, { - "epoch": 0.733386709367494, - "grad_norm": 0.3828125, - "learning_rate": 1.0064390867291774e-05, - "loss": 2.5776, + "epoch": 0.8897307010514558, + "grad_norm": 0.62890625, + "learning_rate": 7.2844390480566126e-06, + "loss": 2.4038, "step": 2290 }, { - "epoch": 0.733706965572458, - "grad_norm": 0.609375, - "learning_rate": 1.00419783321747e-05, - "loss": 2.0832, + "epoch": 0.8901192297418713, + "grad_norm": 0.60546875, + "learning_rate": 7.2336786083926245e-06, + "loss": 2.3153, "step": 2291 }, { - "epoch": 0.734027221777422, - "grad_norm": 0.5546875, - "learning_rate": 1.0019584506910817e-05, - "loss": 1.78, + "epoch": 0.8905077584322867, + "grad_norm": 0.59765625, + "learning_rate": 7.183089008123012e-06, + "loss": 2.2857, "step": 2292 }, { - "epoch": 0.7343474779823859, - "grad_norm": 0.52734375, - "learning_rate": 9.99720941951086e-06, - "loss": 2.3546, + "epoch": 0.8908962871227022, + "grad_norm": 0.609375, + "learning_rate": 7.132670340414105e-06, + "loss": 2.3884, "step": 2293 }, { - "epoch": 0.7346677341873499, - "grad_norm": 0.6328125, - "learning_rate": 9.974853097962116e-06, - "loss": 2.3111, + "epoch": 0.8912848158131177, + "grad_norm": 0.59765625, + "learning_rate": 7.08242269811743e-06, + "loss": 2.2858, "step": 2294 }, { - "epoch": 0.7349879903923139, - "grad_norm": 0.625, - "learning_rate": 9.952515570228432e-06, - "loss": 1.9775, + "epoch": 0.8916733445035332, + "grad_norm": 0.6171875, + "learning_rate": 7.032346173769544e-06, + "loss": 2.345, "step": 2295 }, { - "epoch": 0.7353082465972778, - "grad_norm": 0.494140625, - "learning_rate": 9.930196864250102e-06, - "loss": 2.029, + "epoch": 0.8920618731939487, + "grad_norm": 0.609375, + "learning_rate": 6.98244085959191e-06, + "loss": 2.3408, "step": 2296 }, { - "epoch": 0.7356285028022418, - "grad_norm": 0.49609375, - "learning_rate": 9.90789700794391e-06, - "loss": 2.0413, + "epoch": 0.8924504018843642, + "grad_norm": 0.62890625, + "learning_rate": 6.932706847490622e-06, + "loss": 2.2851, "step": 2297 }, { - "epoch": 0.7359487590072058, - "grad_norm": 0.546875, - "learning_rate": 9.885616029203046e-06, - "loss": 1.8798, + "epoch": 0.8928389305747796, + "grad_norm": 0.58203125, + "learning_rate": 6.8831442290563485e-06, + "loss": 2.3528, "step": 2298 }, { - "epoch": 0.7362690152121697, - "grad_norm": 0.33984375, - "learning_rate": 9.8633539558971e-06, - "loss": 1.7159, + "epoch": 0.8932274592651951, + "grad_norm": 0.60546875, + "learning_rate": 6.833753095564122e-06, + "loss": 2.3832, "step": 2299 }, { - "epoch": 0.7365892714171337, - "grad_norm": 0.466796875, - "learning_rate": 9.84111081587201e-06, - "loss": 2.5267, + "epoch": 0.8936159879556106, + "grad_norm": 0.58984375, + "learning_rate": 6.784533537973137e-06, + "loss": 2.3626, "step": 2300 }, { - "epoch": 0.7369095276220977, - "grad_norm": 1.5234375, - "learning_rate": 9.818886636950037e-06, - "loss": 2.7178, + "epoch": 0.8940045166460261, + "grad_norm": 0.61328125, + "learning_rate": 6.735485646926631e-06, + "loss": 2.3842, "step": 2301 }, { - "epoch": 0.7372297838270616, - "grad_norm": 0.4765625, - "learning_rate": 9.796681446929698e-06, - "loss": 2.4225, + "epoch": 0.8943930453364416, + "grad_norm": 0.6171875, + "learning_rate": 6.686609512751696e-06, + "loss": 2.3222, "step": 2302 }, { - "epoch": 0.7375500400320256, - "grad_norm": 0.4296875, - "learning_rate": 9.77449527358579e-06, - "loss": 2.337, + "epoch": 0.8947815740268571, + "grad_norm": 0.66796875, + "learning_rate": 6.637905225459129e-06, + "loss": 2.4044, "step": 2303 }, { - "epoch": 0.7378702962369896, - "grad_norm": 0.81640625, - "learning_rate": 9.752328144669316e-06, - "loss": 2.5408, + "epoch": 0.8951701027172725, + "grad_norm": 0.62109375, + "learning_rate": 6.589372874743227e-06, + "loss": 2.348, "step": 2304 }, { - "epoch": 0.7381905524419535, - "grad_norm": 0.53515625, - "learning_rate": 9.730180087907445e-06, - "loss": 1.9689, + "epoch": 0.895558631407688, + "grad_norm": 0.5859375, + "learning_rate": 6.541012549981651e-06, + "loss": 2.3778, "step": 2305 }, { - "epoch": 0.7385108086469175, - "grad_norm": 0.396484375, - "learning_rate": 9.708051131003507e-06, - "loss": 2.0099, + "epoch": 0.8959471600981035, + "grad_norm": 0.609375, + "learning_rate": 6.492824340235271e-06, + "loss": 2.3016, "step": 2306 }, { - "epoch": 0.7388310648518815, - "grad_norm": 0.59765625, - "learning_rate": 9.685941301636927e-06, - "loss": 2.1473, + "epoch": 0.896335688788519, + "grad_norm": 0.58203125, + "learning_rate": 6.444808334247965e-06, + "loss": 2.3245, "step": 2307 }, { - "epoch": 0.7391513210568454, - "grad_norm": 0.5625, - "learning_rate": 9.663850627463219e-06, - "loss": 1.9139, + "epoch": 0.8967242174789345, + "grad_norm": 0.62890625, + "learning_rate": 6.396964620446522e-06, + "loss": 2.4092, "step": 2308 }, { - "epoch": 0.7394715772618095, - "grad_norm": 0.59765625, - "learning_rate": 9.641779136113931e-06, - "loss": 2.5305, + "epoch": 0.89711274616935, + "grad_norm": 0.58984375, + "learning_rate": 6.3492932869403835e-06, + "loss": 2.3346, "step": 2309 }, { - "epoch": 0.7397918334667735, - "grad_norm": 0.51171875, - "learning_rate": 9.619726855196603e-06, - "loss": 2.4336, + "epoch": 0.8975012748597654, + "grad_norm": 0.5859375, + "learning_rate": 6.301794421521568e-06, + "loss": 2.3933, "step": 2310 }, { - "epoch": 0.7401120896717374, - "grad_norm": 0.60546875, - "learning_rate": 9.597693812294761e-06, - "loss": 2.3245, + "epoch": 0.8978898035501809, + "grad_norm": 0.57421875, + "learning_rate": 6.25446811166448e-06, + "loss": 2.4146, "step": 2311 }, { - "epoch": 0.7404323458767014, - "grad_norm": 0.5859375, - "learning_rate": 9.57568003496787e-06, - "loss": 2.3801, + "epoch": 0.8982783322405964, + "grad_norm": 0.58203125, + "learning_rate": 6.207314444525703e-06, + "loss": 2.349, "step": 2312 }, { - "epoch": 0.7407526020816654, - "grad_norm": 0.47265625, - "learning_rate": 9.553685550751294e-06, - "loss": 2.4653, + "epoch": 0.8986668609310119, + "grad_norm": 0.6171875, + "learning_rate": 6.160333506943938e-06, + "loss": 2.3358, "step": 2313 }, { - "epoch": 0.7410728582866293, - "grad_norm": 0.494140625, - "learning_rate": 9.531710387156254e-06, - "loss": 2.639, + "epoch": 0.8990553896214274, + "grad_norm": 0.60546875, + "learning_rate": 6.113525385439733e-06, + "loss": 2.3736, "step": 2314 }, { - "epoch": 0.7413931144915933, - "grad_norm": 0.404296875, - "learning_rate": 9.509754571669821e-06, - "loss": 2.0894, + "epoch": 0.8994439183118429, + "grad_norm": 0.6171875, + "learning_rate": 6.066890166215389e-06, + "loss": 2.3436, "step": 2315 }, { - "epoch": 0.7417133706965573, - "grad_norm": 0.53515625, - "learning_rate": 9.48781813175486e-06, - "loss": 1.8972, + "epoch": 0.8998324470022583, + "grad_norm": 0.671875, + "learning_rate": 6.020427935154816e-06, + "loss": 2.3194, "step": 2316 }, { - "epoch": 0.7420336269015212, - "grad_norm": 0.87109375, - "learning_rate": 9.465901094849985e-06, - "loss": 1.8963, + "epoch": 0.9002209756926738, + "grad_norm": 0.640625, + "learning_rate": 5.974138777823312e-06, + "loss": 2.4071, "step": 2317 }, { - "epoch": 0.7423538831064852, - "grad_norm": 0.30078125, - "learning_rate": 9.44400348836956e-06, - "loss": 2.1828, + "epoch": 0.9006095043830893, + "grad_norm": 0.59375, + "learning_rate": 5.928022779467468e-06, + "loss": 2.3346, "step": 2318 }, { - "epoch": 0.7426741393114492, - "grad_norm": 0.578125, - "learning_rate": 9.422125339703635e-06, - "loss": 1.7612, + "epoch": 0.9009980330735048, + "grad_norm": 0.62109375, + "learning_rate": 5.88208002501498e-06, + "loss": 2.2619, "step": 2319 }, { - "epoch": 0.7429943955164131, - "grad_norm": 0.58203125, - "learning_rate": 9.400266676217927e-06, - "loss": 2.154, + "epoch": 0.9013865617639203, + "grad_norm": 0.64453125, + "learning_rate": 5.836310599074468e-06, + "loss": 2.3823, "step": 2320 }, { - "epoch": 0.7433146517213771, - "grad_norm": 0.60546875, - "learning_rate": 9.378427525253781e-06, - "loss": 2.6834, + "epoch": 0.9017750904543358, + "grad_norm": 0.6640625, + "learning_rate": 5.790714585935397e-06, + "loss": 2.3954, "step": 2321 }, { - "epoch": 0.7436349079263411, - "grad_norm": 0.5, - "learning_rate": 9.356607914128126e-06, - "loss": 1.8765, + "epoch": 0.9021636191447512, + "grad_norm": 0.59765625, + "learning_rate": 5.745292069567809e-06, + "loss": 2.366, "step": 2322 }, { - "epoch": 0.743955164131305, - "grad_norm": 0.82421875, - "learning_rate": 9.33480787013346e-06, - "loss": 2.3637, + "epoch": 0.9025521478351667, + "grad_norm": 0.6015625, + "learning_rate": 5.700043133622291e-06, + "loss": 2.4269, "step": 2323 }, { - "epoch": 0.744275420336269, - "grad_norm": 1.203125, - "learning_rate": 9.313027420537809e-06, - "loss": 2.5115, + "epoch": 0.9029406765255822, + "grad_norm": 0.6015625, + "learning_rate": 5.654967861429738e-06, + "loss": 2.3784, "step": 2324 }, { - "epoch": 0.7445956765412329, - "grad_norm": 0.51953125, - "learning_rate": 9.291266592584666e-06, - "loss": 1.9597, + "epoch": 0.9033292052159977, + "grad_norm": 0.6171875, + "learning_rate": 5.61006633600123e-06, + "loss": 2.3596, "step": 2325 }, { - "epoch": 0.7449159327461969, - "grad_norm": 0.466796875, - "learning_rate": 9.269525413493008e-06, - "loss": 2.3502, + "epoch": 0.9037177339064132, + "grad_norm": 0.66015625, + "learning_rate": 5.565338640027862e-06, + "loss": 2.323, "step": 2326 }, { - "epoch": 0.745236188951161, - "grad_norm": 0.44921875, - "learning_rate": 9.247803910457226e-06, - "loss": 2.5435, + "epoch": 0.9041062625968286, + "grad_norm": 0.640625, + "learning_rate": 5.520784855880612e-06, + "loss": 2.3243, "step": 2327 }, { - "epoch": 0.7455564451561248, - "grad_norm": 0.451171875, - "learning_rate": 9.226102110647092e-06, - "loss": 2.3085, + "epoch": 0.9044947912872441, + "grad_norm": 0.59765625, + "learning_rate": 5.4764050656101795e-06, + "loss": 2.3683, "step": 2328 }, { - "epoch": 0.7458767013610889, - "grad_norm": 0.48828125, - "learning_rate": 9.204420041207746e-06, - "loss": 2.0914, + "epoch": 0.9048833199776596, + "grad_norm": 0.58984375, + "learning_rate": 5.432199350946832e-06, + "loss": 2.3238, "step": 2329 }, { - "epoch": 0.7461969575660529, - "grad_norm": 1.6328125, - "learning_rate": 9.182757729259638e-06, - "loss": 1.7195, + "epoch": 0.9052718486680751, + "grad_norm": 0.62109375, + "learning_rate": 5.388167793300281e-06, + "loss": 2.4312, "step": 2330 }, { - "epoch": 0.7465172137710168, - "grad_norm": 0.43359375, - "learning_rate": 9.161115201898518e-06, - "loss": 2.1269, + "epoch": 0.9056603773584906, + "grad_norm": 0.60546875, + "learning_rate": 5.344310473759462e-06, + "loss": 2.32, "step": 2331 }, { - "epoch": 0.7468374699759808, - "grad_norm": 0.609375, - "learning_rate": 9.139492486195364e-06, - "loss": 2.4199, + "epoch": 0.9060489060489061, + "grad_norm": 0.60546875, + "learning_rate": 5.300627473092457e-06, + "loss": 2.3038, "step": 2332 }, { - "epoch": 0.7471577261809448, - "grad_norm": 0.546875, - "learning_rate": 9.117889609196395e-06, - "loss": 2.0749, + "epoch": 0.9064374347393215, + "grad_norm": 0.63671875, + "learning_rate": 5.257118871746347e-06, + "loss": 2.3765, "step": 2333 }, { - "epoch": 0.7474779823859087, - "grad_norm": 0.4921875, - "learning_rate": 9.09630659792301e-06, - "loss": 2.6966, + "epoch": 0.906825963429737, + "grad_norm": 0.61328125, + "learning_rate": 5.213784749846984e-06, + "loss": 2.324, "step": 2334 }, { - "epoch": 0.7477982385908727, - "grad_norm": 0.67578125, - "learning_rate": 9.074743479371761e-06, - "loss": 1.7836, + "epoch": 0.9072144921201525, + "grad_norm": 0.59765625, + "learning_rate": 5.170625187198941e-06, + "loss": 2.3849, "step": 2335 }, { - "epoch": 0.7481184947958367, - "grad_norm": 0.66796875, - "learning_rate": 9.05320028051431e-06, - "loss": 2.27, + "epoch": 0.907603020810568, + "grad_norm": 0.578125, + "learning_rate": 5.127640263285294e-06, + "loss": 2.2514, "step": 2336 }, { - "epoch": 0.7484387510008006, - "grad_norm": 0.57421875, - "learning_rate": 9.031677028297413e-06, - "loss": 1.792, + "epoch": 0.9079915495009835, + "grad_norm": 0.58203125, + "learning_rate": 5.084830057267509e-06, + "loss": 2.3077, "step": 2337 }, { - "epoch": 0.7487590072057646, - "grad_norm": 0.51171875, - "learning_rate": 9.010173749642878e-06, - "loss": 2.2518, + "epoch": 0.908380078191399, + "grad_norm": 0.5859375, + "learning_rate": 5.04219464798531e-06, + "loss": 2.3745, "step": 2338 }, { - "epoch": 0.7490792634107286, - "grad_norm": 0.412109375, - "learning_rate": 8.988690471447503e-06, - "loss": 1.1478, + "epoch": 0.9087686068818144, + "grad_norm": 0.6171875, + "learning_rate": 4.9997341139565e-06, + "loss": 2.35, "step": 2339 }, { - "epoch": 0.7493995196156925, - "grad_norm": 0.49609375, - "learning_rate": 8.967227220583102e-06, - "loss": 2.3766, + "epoch": 0.9091571355722299, + "grad_norm": 0.5859375, + "learning_rate": 4.957448533376819e-06, + "loss": 2.3007, "step": 2340 }, { - "epoch": 0.7497197758206565, - "grad_norm": 0.6015625, - "learning_rate": 8.945784023896425e-06, - "loss": 1.8921, + "epoch": 0.9095456642626454, + "grad_norm": 0.609375, + "learning_rate": 4.915337984119805e-06, + "loss": 2.3123, "step": 2341 }, { - "epoch": 0.7500400320256205, - "grad_norm": 0.486328125, - "learning_rate": 8.924360908209137e-06, - "loss": 1.6513, + "epoch": 0.9099341929530609, + "grad_norm": 0.578125, + "learning_rate": 4.8734025437366826e-06, + "loss": 2.406, "step": 2342 }, { - "epoch": 0.7503602882305844, - "grad_norm": 0.63671875, - "learning_rate": 8.902957900317783e-06, - "loss": 1.5818, + "epoch": 0.9103227216434764, + "grad_norm": 0.59375, + "learning_rate": 4.831642289456184e-06, + "loss": 2.3487, "step": 2343 }, { - "epoch": 0.7506805444355484, - "grad_norm": 0.408203125, - "learning_rate": 8.881575026993766e-06, - "loss": 2.2554, + "epoch": 0.9107112503338919, + "grad_norm": 0.6015625, + "learning_rate": 4.790057298184425e-06, + "loss": 2.3409, "step": 2344 }, { - "epoch": 0.7510008006405124, - "grad_norm": 0.56640625, - "learning_rate": 8.860212314983292e-06, - "loss": 2.2229, + "epoch": 0.9110997790243073, + "grad_norm": 0.6171875, + "learning_rate": 4.748647646504722e-06, + "loss": 2.3852, "step": 2345 }, { - "epoch": 0.7513210568454763, - "grad_norm": 1.6484375, - "learning_rate": 8.838869791007367e-06, - "loss": 2.2565, + "epoch": 0.9114883077147228, + "grad_norm": 0.6171875, + "learning_rate": 4.70741341067753e-06, + "loss": 2.3947, "step": 2346 }, { - "epoch": 0.7516413130504404, - "grad_norm": 0.5078125, - "learning_rate": 8.817547481761717e-06, - "loss": 2.5122, + "epoch": 0.9118768364051383, + "grad_norm": 0.640625, + "learning_rate": 4.666354666640216e-06, + "loss": 2.4049, "step": 2347 }, { - "epoch": 0.7519615692554044, - "grad_norm": 0.47265625, - "learning_rate": 8.796245413916807e-06, - "loss": 2.6242, + "epoch": 0.9122653650955538, + "grad_norm": 0.6015625, + "learning_rate": 4.625471490007005e-06, + "loss": 2.3076, "step": 2348 }, { - "epoch": 0.7522818254603683, - "grad_norm": 0.60546875, - "learning_rate": 8.774963614117773e-06, - "loss": 2.2074, + "epoch": 0.9126538937859693, + "grad_norm": 0.62890625, + "learning_rate": 4.584763956068738e-06, + "loss": 2.3434, "step": 2349 }, { - "epoch": 0.7526020816653323, - "grad_norm": 0.56640625, - "learning_rate": 8.753702108984405e-06, - "loss": 2.5946, + "epoch": 0.9130424224763848, + "grad_norm": 0.6328125, + "learning_rate": 4.544232139792826e-06, + "loss": 2.325, "step": 2350 }, { - "epoch": 0.7529223378702963, - "grad_norm": 0.431640625, - "learning_rate": 8.732460925111105e-06, - "loss": 2.4789, + "epoch": 0.9134309511668002, + "grad_norm": 0.6484375, + "learning_rate": 4.503876115823081e-06, + "loss": 2.4135, "step": 2351 }, { - "epoch": 0.7532425940752602, - "grad_norm": 0.58203125, - "learning_rate": 8.711240089066863e-06, - "loss": 2.1403, + "epoch": 0.9138194798572157, + "grad_norm": 0.640625, + "learning_rate": 4.463695958479563e-06, + "loss": 2.4263, "step": 2352 }, { - "epoch": 0.7535628502802242, - "grad_norm": 0.80078125, - "learning_rate": 8.690039627395195e-06, - "loss": 2.0411, + "epoch": 0.9142080085476312, + "grad_norm": 0.6171875, + "learning_rate": 4.423691741758451e-06, + "loss": 2.3761, "step": 2353 }, { - "epoch": 0.7538831064851882, - "grad_norm": 0.494140625, - "learning_rate": 8.668859566614154e-06, - "loss": 2.4965, + "epoch": 0.9145965372380467, + "grad_norm": 0.60546875, + "learning_rate": 4.383863539331923e-06, + "loss": 2.3851, "step": 2354 }, { - "epoch": 0.7542033626901521, - "grad_norm": 0.54296875, - "learning_rate": 8.647699933216274e-06, - "loss": 2.275, + "epoch": 0.9149850659284622, + "grad_norm": 0.5703125, + "learning_rate": 4.3442114245479835e-06, + "loss": 2.2704, "step": 2355 }, { - "epoch": 0.7545236188951161, - "grad_norm": 0.55078125, - "learning_rate": 8.626560753668531e-06, - "loss": 2.3405, + "epoch": 0.9153735946188777, + "grad_norm": 0.640625, + "learning_rate": 4.304735470430387e-06, + "loss": 2.3783, "step": 2356 }, { - "epoch": 0.7548438751000801, - "grad_norm": 0.498046875, - "learning_rate": 8.605442054412318e-06, - "loss": 2.0917, + "epoch": 0.9157621233092931, + "grad_norm": 0.6171875, + "learning_rate": 4.265435749678448e-06, + "loss": 2.4151, "step": 2357 }, { - "epoch": 0.755164131305044, - "grad_norm": 0.515625, - "learning_rate": 8.584343861863418e-06, - "loss": 2.1474, + "epoch": 0.9161506519997086, + "grad_norm": 0.5859375, + "learning_rate": 4.226312334666904e-06, + "loss": 2.2984, "step": 2358 }, { - "epoch": 0.755484387510008, - "grad_norm": 0.4765625, - "learning_rate": 8.563266202411948e-06, - "loss": 2.2758, + "epoch": 0.9165391806901241, + "grad_norm": 0.58203125, + "learning_rate": 4.187365297445844e-06, + "loss": 2.3461, "step": 2359 }, { - "epoch": 0.755804643714972, - "grad_norm": 0.59765625, - "learning_rate": 8.542209102422366e-06, - "loss": 2.5798, + "epoch": 0.9169277093805396, + "grad_norm": 0.609375, + "learning_rate": 4.148594709740539e-06, + "loss": 2.3782, "step": 2360 }, { - "epoch": 0.7561248999199359, - "grad_norm": 0.5546875, - "learning_rate": 8.521172588233378e-06, - "loss": 2.485, + "epoch": 0.9173162380709551, + "grad_norm": 0.6015625, + "learning_rate": 4.110000642951273e-06, + "loss": 2.3375, "step": 2361 }, { - "epoch": 0.7564451561248999, - "grad_norm": 0.5, - "learning_rate": 8.500156686157975e-06, - "loss": 2.354, + "epoch": 0.9177047667613706, + "grad_norm": 0.58203125, + "learning_rate": 4.071583168153293e-06, + "loss": 2.3792, "step": 2362 }, { - "epoch": 0.7567654123298639, - "grad_norm": 0.47265625, - "learning_rate": 8.479161422483353e-06, - "loss": 1.9641, + "epoch": 0.918093295451786, + "grad_norm": 0.6015625, + "learning_rate": 4.033342356096592e-06, + "loss": 2.3389, "step": 2363 }, { - "epoch": 0.7570856685348278, - "grad_norm": 0.43359375, - "learning_rate": 8.458186823470888e-06, - "loss": 2.5159, + "epoch": 0.9184818241422015, + "grad_norm": 0.62890625, + "learning_rate": 3.995278277205839e-06, + "loss": 2.4245, "step": 2364 }, { - "epoch": 0.7574059247397918, - "grad_norm": 0.50390625, - "learning_rate": 8.437232915356114e-06, - "loss": 2.1078, + "epoch": 0.918870352832617, + "grad_norm": 0.6015625, + "learning_rate": 3.957391001580235e-06, + "loss": 2.2686, "step": 2365 }, { - "epoch": 0.7577261809447559, - "grad_norm": 0.59765625, - "learning_rate": 8.416299724348695e-06, - "loss": 2.3243, + "epoch": 0.9192588815230325, + "grad_norm": 0.60546875, + "learning_rate": 3.919680598993347e-06, + "loss": 2.3117, "step": 2366 }, { - "epoch": 0.7580464371497198, - "grad_norm": 0.515625, - "learning_rate": 8.395387276632338e-06, - "loss": 1.7567, + "epoch": 0.919647410213448, + "grad_norm": 0.58984375, + "learning_rate": 3.882147138893055e-06, + "loss": 2.3161, "step": 2367 }, { - "epoch": 0.7583666933546838, - "grad_norm": 0.44921875, - "learning_rate": 8.374495598364876e-06, - "loss": 2.1981, + "epoch": 0.9200359389038635, + "grad_norm": 0.62890625, + "learning_rate": 3.844790690401356e-06, + "loss": 2.3478, "step": 2368 }, { - "epoch": 0.7586869495596478, - "grad_norm": 0.609375, - "learning_rate": 8.353624715678093e-06, - "loss": 2.1722, + "epoch": 0.9204244675942789, + "grad_norm": 0.6484375, + "learning_rate": 3.807611322314242e-06, + "loss": 2.303, "step": 2369 }, { - "epoch": 0.7590072057646117, - "grad_norm": 0.53515625, - "learning_rate": 8.332774654677802e-06, - "loss": 2.2055, + "epoch": 0.9208129962846944, + "grad_norm": 0.63671875, + "learning_rate": 3.770609103101641e-06, + "loss": 2.4052, "step": 2370 }, { - "epoch": 0.7593274619695757, - "grad_norm": 0.55078125, - "learning_rate": 8.311945441443759e-06, - "loss": 2.5394, + "epoch": 0.9212015249751099, + "grad_norm": 0.62890625, + "learning_rate": 3.7337841009072007e-06, + "loss": 2.3042, "step": 2371 }, { - "epoch": 0.7596477181745396, - "grad_norm": 0.5234375, - "learning_rate": 8.291137102029644e-06, - "loss": 2.7369, + "epoch": 0.9215900536655254, + "grad_norm": 0.62109375, + "learning_rate": 3.6971363835482163e-06, + "loss": 2.3081, "step": 2372 }, { - "epoch": 0.7599679743795036, - "grad_norm": 0.4453125, - "learning_rate": 8.270349662463032e-06, - "loss": 2.2094, + "epoch": 0.9219785823559409, + "grad_norm": 0.58984375, + "learning_rate": 3.660666018515491e-06, + "loss": 2.3679, "step": 2373 }, { - "epoch": 0.7602882305844676, - "grad_norm": 0.373046875, - "learning_rate": 8.249583148745357e-06, - "loss": 2.7189, + "epoch": 0.9223671110463564, + "grad_norm": 0.61328125, + "learning_rate": 3.624373072973242e-06, + "loss": 2.2973, "step": 2374 }, { - "epoch": 0.7606084867894315, - "grad_norm": 0.51953125, - "learning_rate": 8.228837586851859e-06, - "loss": 2.4148, + "epoch": 0.9227556397367718, + "grad_norm": 0.625, + "learning_rate": 3.588257613758883e-06, + "loss": 2.4296, "step": 2375 }, { - "epoch": 0.7609287429943955, - "grad_norm": 0.58203125, - "learning_rate": 8.20811300273159e-06, - "loss": 2.1258, + "epoch": 0.9231441684271873, + "grad_norm": 0.60546875, + "learning_rate": 3.5523197073830337e-06, + "loss": 2.2725, "step": 2376 }, { - "epoch": 0.7612489991993595, - "grad_norm": 0.52734375, - "learning_rate": 8.187409422307365e-06, - "loss": 1.9111, + "epoch": 0.9235326971176028, + "grad_norm": 0.5859375, + "learning_rate": 3.5165594200293193e-06, + "loss": 2.3307, "step": 2377 }, { - "epoch": 0.7615692554043234, - "grad_norm": 0.427734375, - "learning_rate": 8.166726871475713e-06, - "loss": 2.2352, + "epoch": 0.9239212258080183, + "grad_norm": 0.625, + "learning_rate": 3.4809768175542046e-06, + "loss": 2.3393, "step": 2378 }, { - "epoch": 0.7618895116092874, - "grad_norm": 0.546875, - "learning_rate": 8.146065376106873e-06, - "loss": 2.1225, + "epoch": 0.9243097544984338, + "grad_norm": 0.59765625, + "learning_rate": 3.4455719654870045e-06, + "loss": 2.3512, "step": 2379 }, { - "epoch": 0.7622097678142514, - "grad_norm": 1.0078125, - "learning_rate": 8.125424962044742e-06, - "loss": 2.1848, + "epoch": 0.9246982831888493, + "grad_norm": 0.58984375, + "learning_rate": 3.4103449290296297e-06, + "loss": 2.3553, "step": 2380 }, { - "epoch": 0.7625300240192153, - "grad_norm": 0.6015625, - "learning_rate": 8.10480565510683e-06, - "loss": 2.2108, + "epoch": 0.9250868118792647, + "grad_norm": 0.59375, + "learning_rate": 3.375295773056564e-06, + "loss": 2.3058, "step": 2381 }, { - "epoch": 0.7628502802241793, - "grad_norm": 0.484375, - "learning_rate": 8.084207481084282e-06, - "loss": 2.359, + "epoch": 0.9254753405696802, + "grad_norm": 0.609375, + "learning_rate": 3.3404245621146855e-06, + "loss": 2.3938, "step": 2382 }, { - "epoch": 0.7631705364291433, - "grad_norm": 0.49609375, - "learning_rate": 8.063630465741775e-06, - "loss": 1.9673, + "epoch": 0.9258638692600957, + "grad_norm": 0.63671875, + "learning_rate": 3.305731360423159e-06, + "loss": 2.3489, "step": 2383 }, { - "epoch": 0.7634907926341072, - "grad_norm": 0.609375, - "learning_rate": 8.043074634817541e-06, - "loss": 2.294, + "epoch": 0.9262523979505112, + "grad_norm": 0.58984375, + "learning_rate": 3.271216231873353e-06, + "loss": 2.3626, "step": 2384 }, { - "epoch": 0.7638110488390712, - "grad_norm": 0.404296875, - "learning_rate": 8.022540014023306e-06, - "loss": 2.1543, + "epoch": 0.9266409266409267, + "grad_norm": 0.5703125, + "learning_rate": 3.2368792400286453e-06, + "loss": 2.3769, "step": 2385 }, { - "epoch": 0.7641313050440353, - "grad_norm": 0.474609375, - "learning_rate": 8.002026629044268e-06, - "loss": 2.2717, + "epoch": 0.9270294553313422, + "grad_norm": 0.61328125, + "learning_rate": 3.202720448124408e-06, + "loss": 2.2995, "step": 2386 }, { - "epoch": 0.7644515612489992, - "grad_norm": 0.4609375, - "learning_rate": 7.981534505539063e-06, - "loss": 2.7924, + "epoch": 0.9274179840217576, + "grad_norm": 0.6015625, + "learning_rate": 3.16873991906782e-06, + "loss": 2.3558, "step": 2387 }, { - "epoch": 0.7647718174539632, - "grad_norm": 0.466796875, - "learning_rate": 7.961063669139738e-06, - "loss": 2.4514, + "epoch": 0.9278065127121731, + "grad_norm": 0.56640625, + "learning_rate": 3.134937715437758e-06, + "loss": 2.2806, "step": 2388 }, { - "epoch": 0.7650920736589272, - "grad_norm": 0.4609375, - "learning_rate": 7.940614145451683e-06, - "loss": 2.3244, + "epoch": 0.9281950414025886, + "grad_norm": 0.62109375, + "learning_rate": 3.1013138994847036e-06, + "loss": 2.3126, "step": 2389 }, { - "epoch": 0.7654123298638911, - "grad_norm": 1.03125, - "learning_rate": 7.920185960053688e-06, - "loss": 1.8777, + "epoch": 0.9285835700930041, + "grad_norm": 0.59765625, + "learning_rate": 3.0678685331306133e-06, + "loss": 2.3287, "step": 2390 }, { - "epoch": 0.7657325860688551, - "grad_norm": 0.5859375, - "learning_rate": 7.899779138497796e-06, - "loss": 2.4442, + "epoch": 0.9289720987834196, + "grad_norm": 0.62890625, + "learning_rate": 3.034601677968818e-06, + "loss": 2.3597, "step": 2391 }, { - "epoch": 0.7660528422738191, - "grad_norm": 0.9609375, - "learning_rate": 7.879393706309354e-06, - "loss": 2.174, + "epoch": 0.929360627473835, + "grad_norm": 0.60546875, + "learning_rate": 3.0015133952638994e-06, + "loss": 2.3254, "step": 2392 }, { - "epoch": 0.766373098478783, - "grad_norm": 0.6015625, - "learning_rate": 7.859029688986952e-06, - "loss": 1.5029, + "epoch": 0.9297491561642505, + "grad_norm": 0.65234375, + "learning_rate": 2.96860374595157e-06, + "loss": 2.3784, "step": 2393 }, { - "epoch": 0.766693354683747, - "grad_norm": 0.54296875, - "learning_rate": 7.838687112002397e-06, - "loss": 2.0569, + "epoch": 0.930137684854666, + "grad_norm": 0.62109375, + "learning_rate": 2.9358727906385607e-06, + "loss": 2.3332, "step": 2394 }, { - "epoch": 0.767013610888711, - "grad_norm": 0.5078125, - "learning_rate": 7.818366000800666e-06, - "loss": 2.0224, + "epoch": 0.9305262135450815, + "grad_norm": 0.609375, + "learning_rate": 2.9033205896025316e-06, + "loss": 2.338, "step": 2395 }, { - "epoch": 0.7673338670936749, - "grad_norm": 0.474609375, - "learning_rate": 7.798066380799907e-06, - "loss": 2.3784, + "epoch": 0.930914742235497, + "grad_norm": 0.61328125, + "learning_rate": 2.8709472027919405e-06, + "loss": 2.3347, "step": 2396 }, { - "epoch": 0.7676541232986389, - "grad_norm": 0.478515625, - "learning_rate": 7.777788277391348e-06, - "loss": 2.0502, + "epoch": 0.9313032709259125, + "grad_norm": 0.609375, + "learning_rate": 2.838752689825963e-06, + "loss": 2.3698, "step": 2397 }, { - "epoch": 0.7679743795036029, - "grad_norm": 0.455078125, - "learning_rate": 7.757531715939358e-06, - "loss": 1.6753, + "epoch": 0.9316917996163279, + "grad_norm": 0.578125, + "learning_rate": 2.8067371099943286e-06, + "loss": 2.4252, "step": 2398 }, { - "epoch": 0.7682946357085668, - "grad_norm": 0.515625, - "learning_rate": 7.73729672178131e-06, - "loss": 2.408, + "epoch": 0.9320803283067434, + "grad_norm": 0.62109375, + "learning_rate": 2.774900522257251e-06, + "loss": 2.3157, "step": 2399 }, { - "epoch": 0.7686148919135308, - "grad_norm": 0.48046875, - "learning_rate": 7.717083320227627e-06, - "loss": 2.0858, + "epoch": 0.9324688569971589, + "grad_norm": 0.6484375, + "learning_rate": 2.74324298524532e-06, + "loss": 2.341, "step": 2400 }, { - "epoch": 0.7689351481184948, - "grad_norm": 0.59765625, - "learning_rate": 7.69689153656172e-06, - "loss": 2.4706, + "epoch": 0.9328573856875744, + "grad_norm": 0.578125, + "learning_rate": 2.7117645572593777e-06, + "loss": 2.3027, "step": 2401 }, { - "epoch": 0.7692554043234587, - "grad_norm": 1.125, - "learning_rate": 7.676721396039963e-06, - "loss": 2.2135, + "epoch": 0.9332459143779899, + "grad_norm": 0.625, + "learning_rate": 2.6804652962703847e-06, + "loss": 2.4317, "step": 2402 }, { - "epoch": 0.7695756605284227, - "grad_norm": 0.482421875, - "learning_rate": 7.65657292389163e-06, - "loss": 2.5483, + "epoch": 0.9336344430684054, + "grad_norm": 0.60546875, + "learning_rate": 2.6493452599194112e-06, + "loss": 2.4147, "step": 2403 }, { - "epoch": 0.7698959167333868, - "grad_norm": 0.515625, - "learning_rate": 7.636446145318945e-06, - "loss": 2.4565, + "epoch": 0.9340229717588208, + "grad_norm": 0.6171875, + "learning_rate": 2.6184045055174024e-06, + "loss": 2.3043, "step": 2404 }, { - "epoch": 0.7702161729383507, - "grad_norm": 0.51171875, - "learning_rate": 7.6163410854969476e-06, - "loss": 2.0011, + "epoch": 0.9344115004492363, + "grad_norm": 0.609375, + "learning_rate": 2.5876430900451886e-06, + "loss": 2.3907, "step": 2405 }, { - "epoch": 0.7705364291433147, - "grad_norm": 0.55078125, - "learning_rate": 7.596257769573534e-06, - "loss": 2.4179, + "epoch": 0.9348000291396518, + "grad_norm": 0.640625, + "learning_rate": 2.5570610701532995e-06, + "loss": 2.3827, "step": 2406 }, { - "epoch": 0.7708566853482787, - "grad_norm": 1.046875, - "learning_rate": 7.576196222669396e-06, - "loss": 2.2249, + "epoch": 0.9351885578300673, + "grad_norm": 0.5859375, + "learning_rate": 2.5266585021618717e-06, + "loss": 2.4008, "step": 2407 }, { - "epoch": 0.7711769415532426, - "grad_norm": 0.55859375, - "learning_rate": 7.5561564698780065e-06, - "loss": 2.3321, + "epoch": 0.9355770865204828, + "grad_norm": 0.6015625, + "learning_rate": 2.4964354420606073e-06, + "loss": 2.375, "step": 2408 }, { - "epoch": 0.7714971977582066, - "grad_norm": 0.79296875, - "learning_rate": 7.536138536265563e-06, - "loss": 2.2056, + "epoch": 0.9359656152108983, + "grad_norm": 0.640625, + "learning_rate": 2.4663919455085727e-06, + "loss": 2.3972, "step": 2409 }, { - "epoch": 0.7718174539631706, - "grad_norm": 0.5703125, - "learning_rate": 7.516142446870989e-06, - "loss": 2.3361, + "epoch": 0.9363541439013137, + "grad_norm": 0.6015625, + "learning_rate": 2.4365280678342094e-06, + "loss": 2.3483, "step": 2410 }, { - "epoch": 0.7721377101681345, - "grad_norm": 0.4765625, - "learning_rate": 7.496168226705852e-06, - "loss": 1.8415, + "epoch": 0.9367426725917292, + "grad_norm": 0.625, + "learning_rate": 2.406843864035102e-06, + "loss": 2.3772, "step": 2411 }, { - "epoch": 0.7724579663730985, - "grad_norm": 0.5546875, - "learning_rate": 7.476215900754413e-06, - "loss": 2.214, + "epoch": 0.9371312012821447, + "grad_norm": 0.609375, + "learning_rate": 2.377339388777999e-06, + "loss": 2.3935, "step": 2412 }, { - "epoch": 0.7727782225780625, - "grad_norm": 0.435546875, - "learning_rate": 7.456285493973506e-06, - "loss": 2.1685, + "epoch": 0.9375197299725602, + "grad_norm": 0.61328125, + "learning_rate": 2.348014696398626e-06, + "loss": 2.3582, "step": 2413 }, { - "epoch": 0.7730984787830264, - "grad_norm": 0.5078125, - "learning_rate": 7.4363770312925625e-06, - "loss": 1.6529, + "epoch": 0.9379082586629757, + "grad_norm": 0.625, + "learning_rate": 2.3188698409016494e-06, + "loss": 2.4331, "step": 2414 }, { - "epoch": 0.7734187349879904, - "grad_norm": 0.5625, - "learning_rate": 7.416490537613571e-06, - "loss": 2.5159, + "epoch": 0.9382967873533912, + "grad_norm": 0.5859375, + "learning_rate": 2.2899048759605136e-06, + "loss": 2.3858, "step": 2415 }, { - "epoch": 0.7737389911929544, - "grad_norm": 0.5390625, - "learning_rate": 7.396626037811042e-06, - "loss": 1.9631, + "epoch": 0.9386853160438066, + "grad_norm": 0.61328125, + "learning_rate": 2.261119854917404e-06, + "loss": 2.3011, "step": 2416 }, { - "epoch": 0.7740592473979183, - "grad_norm": 0.67578125, - "learning_rate": 7.376783556731948e-06, - "loss": 2.3239, + "epoch": 0.9390738447342221, + "grad_norm": 0.60546875, + "learning_rate": 2.2325148307831057e-06, + "loss": 2.4114, "step": 2417 }, { - "epoch": 0.7743795036028823, - "grad_norm": 0.478515625, - "learning_rate": 7.35696311919577e-06, - "loss": 1.8956, + "epoch": 0.9394623734246376, + "grad_norm": 0.65234375, + "learning_rate": 2.204089856236913e-06, + "loss": 2.4188, "step": 2418 }, { - "epoch": 0.7746997598078462, - "grad_norm": 0.91015625, - "learning_rate": 7.337164749994368e-06, - "loss": 2.2789, + "epoch": 0.9398509021150531, + "grad_norm": 0.6015625, + "learning_rate": 2.1758449836265403e-06, + "loss": 2.2944, "step": 2419 }, { - "epoch": 0.7750200160128102, - "grad_norm": 0.4921875, - "learning_rate": 7.317388473892023e-06, - "loss": 2.1517, + "epoch": 0.9402394308054686, + "grad_norm": 0.6015625, + "learning_rate": 2.147780264968047e-06, + "loss": 2.3652, "step": 2420 }, { - "epoch": 0.7753402722177742, - "grad_norm": 0.40234375, - "learning_rate": 7.297634315625377e-06, - "loss": 1.8434, + "epoch": 0.9406279594958841, + "grad_norm": 0.62890625, + "learning_rate": 2.1198957519456887e-06, + "loss": 2.3548, "step": 2421 }, { - "epoch": 0.7756605284227381, - "grad_norm": 0.462890625, - "learning_rate": 7.277902299903405e-06, - "loss": 2.4362, + "epoch": 0.9410164881862995, + "grad_norm": 0.6171875, + "learning_rate": 2.092191495911877e-06, + "loss": 2.3325, "step": 2422 }, { - "epoch": 0.7759807846277021, - "grad_norm": 0.478515625, - "learning_rate": 7.2581924514073916e-06, - "loss": 2.2693, + "epoch": 0.941405016876715, + "grad_norm": 0.60546875, + "learning_rate": 2.0646675478870337e-06, + "loss": 2.3293, "step": 2423 }, { - "epoch": 0.7763010408326662, - "grad_norm": 0.64453125, - "learning_rate": 7.238504794790893e-06, - "loss": 2.4455, + "epoch": 0.9417935455671305, + "grad_norm": 0.60546875, + "learning_rate": 2.037323958559545e-06, + "loss": 2.4153, "step": 2424 }, { - "epoch": 0.77662129703763, - "grad_norm": 0.515625, - "learning_rate": 7.218839354679683e-06, - "loss": 2.5956, + "epoch": 0.942182074257546, + "grad_norm": 0.6015625, + "learning_rate": 2.0101607782856526e-06, + "loss": 2.306, "step": 2425 }, { - "epoch": 0.7769415532425941, - "grad_norm": 0.86328125, - "learning_rate": 7.199196155671795e-06, - "loss": 2.187, + "epoch": 0.9425706029479615, + "grad_norm": 0.6015625, + "learning_rate": 1.983178057089341e-06, + "loss": 2.3611, "step": 2426 }, { - "epoch": 0.7772618094475581, - "grad_norm": 0.4921875, - "learning_rate": 7.179575222337393e-06, - "loss": 2.5182, + "epoch": 0.942959131638377, + "grad_norm": 0.64453125, + "learning_rate": 1.9563758446622502e-06, + "loss": 2.3424, "step": 2427 }, { - "epoch": 0.777582065652522, - "grad_norm": 0.84765625, - "learning_rate": 7.159976579218819e-06, - "loss": 1.9639, + "epoch": 0.9433476603287924, + "grad_norm": 0.61328125, + "learning_rate": 1.9297541903636196e-06, + "loss": 2.3627, "step": 2428 }, { - "epoch": 0.777902321857486, - "grad_norm": 0.53515625, - "learning_rate": 7.140400250830529e-06, - "loss": 2.2367, + "epoch": 0.9437361890192079, + "grad_norm": 0.62890625, + "learning_rate": 1.9033131432201424e-06, + "loss": 2.349, "step": 2429 }, { - "epoch": 0.77822257806245, - "grad_norm": 0.58984375, - "learning_rate": 7.12084626165907e-06, - "loss": 2.4143, + "epoch": 0.9441247177096234, + "grad_norm": 0.6484375, + "learning_rate": 1.877052751925934e-06, + "loss": 2.3583, "step": 2430 }, { - "epoch": 0.7785428342674139, - "grad_norm": 0.55859375, - "learning_rate": 7.10131463616302e-06, - "loss": 1.9487, + "epoch": 0.9445132464000389, + "grad_norm": 0.66796875, + "learning_rate": 1.8509730648423762e-06, + "loss": 2.2995, "step": 2431 }, { - "epoch": 0.7788630904723779, - "grad_norm": 0.64453125, - "learning_rate": 7.0818053987730395e-06, - "loss": 2.0972, + "epoch": 0.9449017750904544, + "grad_norm": 0.63671875, + "learning_rate": 1.8250741299980945e-06, + "loss": 2.331, "step": 2432 }, { - "epoch": 0.7791833466773419, - "grad_norm": 0.5234375, - "learning_rate": 7.062318573891716e-06, - "loss": 2.4615, + "epoch": 0.9452903037808699, + "grad_norm": 0.6015625, + "learning_rate": 1.7993559950888362e-06, + "loss": 2.3712, "step": 2433 }, { - "epoch": 0.7795036028823058, - "grad_norm": 0.55859375, - "learning_rate": 7.042854185893674e-06, - "loss": 2.0897, + "epoch": 0.9456788324712853, + "grad_norm": 0.59765625, + "learning_rate": 1.7738187074773705e-06, + "loss": 2.3326, "step": 2434 }, { - "epoch": 0.7798238590872698, - "grad_norm": 0.83984375, - "learning_rate": 7.023412259125414e-06, - "loss": 2.435, + "epoch": 0.9460673611617008, + "grad_norm": 0.6171875, + "learning_rate": 1.748462314193422e-06, + "loss": 2.3624, "step": 2435 }, { - "epoch": 0.7801441152922338, - "grad_norm": 0.451171875, - "learning_rate": 7.00399281790537e-06, - "loss": 2.3402, + "epoch": 0.9464558898521163, + "grad_norm": 0.6015625, + "learning_rate": 1.723286861933593e-06, + "loss": 2.4015, "step": 2436 }, { - "epoch": 0.7804643714971977, - "grad_norm": 0.5234375, - "learning_rate": 6.984595886523848e-06, - "loss": 2.1758, + "epoch": 0.9468444185425318, + "grad_norm": 0.5859375, + "learning_rate": 1.6982923970612409e-06, + "loss": 2.3807, "step": 2437 }, { - "epoch": 0.7807846277021617, - "grad_norm": 0.416015625, - "learning_rate": 6.965221489243007e-06, - "loss": 2.3811, + "epoch": 0.9472329472329473, + "grad_norm": 0.625, + "learning_rate": 1.673478965606423e-06, + "loss": 2.3904, "step": 2438 }, { - "epoch": 0.7811048839071257, - "grad_norm": 0.337890625, - "learning_rate": 6.94586965029678e-06, - "loss": 1.8303, + "epoch": 0.9476214759233628, + "grad_norm": 0.609375, + "learning_rate": 1.6488466132658087e-06, + "loss": 2.327, "step": 2439 }, { - "epoch": 0.7814251401120896, - "grad_norm": 0.53515625, - "learning_rate": 6.926540393890948e-06, - "loss": 2.0528, + "epoch": 0.9480100046137782, + "grad_norm": 0.578125, + "learning_rate": 1.6243953854025773e-06, + "loss": 2.3492, "step": 2440 }, { - "epoch": 0.7817453963170536, - "grad_norm": 0.3828125, - "learning_rate": 6.907233744202985e-06, - "loss": 2.1735, + "epoch": 0.9483985333041937, + "grad_norm": 0.609375, + "learning_rate": 1.6001253270463757e-06, + "loss": 2.3737, "step": 2441 }, { - "epoch": 0.7820656525220177, - "grad_norm": 0.51953125, - "learning_rate": 6.887949725382117e-06, - "loss": 2.369, + "epoch": 0.9487870619946092, + "grad_norm": 0.61328125, + "learning_rate": 1.5760364828931728e-06, + "loss": 2.3146, "step": 2442 }, { - "epoch": 0.7823859087269815, - "grad_norm": 0.5703125, - "learning_rate": 6.8686883615492666e-06, - "loss": 2.4711, + "epoch": 0.9491755906850247, + "grad_norm": 0.59375, + "learning_rate": 1.5521288973052273e-06, + "loss": 2.3018, "step": 2443 }, { - "epoch": 0.7827061649319456, - "grad_norm": 0.478515625, - "learning_rate": 6.84944967679701e-06, - "loss": 2.5019, + "epoch": 0.9495641193754402, + "grad_norm": 0.5859375, + "learning_rate": 1.5284026143110087e-06, + "loss": 2.3319, "step": 2444 }, { - "epoch": 0.7830264211369096, - "grad_norm": 0.55859375, - "learning_rate": 6.830233695189539e-06, - "loss": 2.4643, + "epoch": 0.9499526480658557, + "grad_norm": 0.58984375, + "learning_rate": 1.504857677605065e-06, + "loss": 2.3617, "step": 2445 }, { - "epoch": 0.7833466773418735, - "grad_norm": 0.67578125, - "learning_rate": 6.8110404407626946e-06, - "loss": 1.7722, + "epoch": 0.950341176756271, + "grad_norm": 0.59375, + "learning_rate": 1.481494130547989e-06, + "loss": 2.3314, "step": 2446 }, { - "epoch": 0.7836669335468375, - "grad_norm": 0.365234375, - "learning_rate": 6.791869937523831e-06, - "loss": 1.7393, + "epoch": 0.9507297054466866, + "grad_norm": 0.58203125, + "learning_rate": 1.45831201616633e-06, + "loss": 2.3293, "step": 2447 }, { - "epoch": 0.7839871897518015, - "grad_norm": 0.396484375, - "learning_rate": 6.772722209451904e-06, - "loss": 1.9059, + "epoch": 0.9511182341371021, + "grad_norm": 0.58984375, + "learning_rate": 1.435311377152493e-06, + "loss": 2.3415, "step": 2448 }, { - "epoch": 0.7843074459567654, - "grad_norm": 0.57421875, - "learning_rate": 6.753597280497329e-06, - "loss": 1.9944, + "epoch": 0.9515067628275176, + "grad_norm": 0.5859375, + "learning_rate": 1.4124922558646946e-06, + "loss": 2.3067, "step": 2449 }, { - "epoch": 0.7846277021617294, - "grad_norm": 0.44921875, - "learning_rate": 6.734495174582037e-06, - "loss": 2.2997, + "epoch": 0.9518952915179331, + "grad_norm": 0.61328125, + "learning_rate": 1.3898546943268643e-06, + "loss": 2.4032, "step": 2450 }, { - "epoch": 0.7849479583666934, - "grad_norm": 0.392578125, - "learning_rate": 6.715415915599399e-06, - "loss": 2.5738, + "epoch": 0.9522838202083486, + "grad_norm": 0.65625, + "learning_rate": 1.3673987342285533e-06, + "loss": 2.3676, "step": 2451 }, { - "epoch": 0.7852682145716573, - "grad_norm": 0.474609375, - "learning_rate": 6.69635952741422e-06, - "loss": 2.5298, + "epoch": 0.952672348898764, + "grad_norm": 0.6015625, + "learning_rate": 1.3451244169249033e-06, + "loss": 2.3903, "step": 2452 }, { - "epoch": 0.7855884707766213, - "grad_norm": 0.341796875, - "learning_rate": 6.677326033862666e-06, - "loss": 2.172, + "epoch": 0.9530608775891795, + "grad_norm": 0.6328125, + "learning_rate": 1.3230317834365013e-06, + "loss": 2.367, "step": 2453 }, { - "epoch": 0.7859087269815853, - "grad_norm": 0.37109375, - "learning_rate": 6.658315458752315e-06, - "loss": 1.9617, + "epoch": 0.953449406279595, + "grad_norm": 0.6015625, + "learning_rate": 1.3011208744493796e-06, + "loss": 2.3764, "step": 2454 }, { - "epoch": 0.7862289831865492, - "grad_norm": 0.5234375, - "learning_rate": 6.639327825862035e-06, - "loss": 2.0634, + "epoch": 0.9538379349700105, + "grad_norm": 0.66796875, + "learning_rate": 1.2793917303148716e-06, + "loss": 2.3566, "step": 2455 }, { - "epoch": 0.7865492393915132, - "grad_norm": 0.5078125, - "learning_rate": 6.620363158942016e-06, - "loss": 2.2057, + "epoch": 0.954226463660426, + "grad_norm": 0.78125, + "learning_rate": 1.2578443910496008e-06, + "loss": 2.4127, "step": 2456 }, { - "epoch": 0.7868694955964772, - "grad_norm": 0.5234375, - "learning_rate": 6.601421481713724e-06, - "loss": 2.4316, + "epoch": 0.9546149923508415, + "grad_norm": 0.69140625, + "learning_rate": 1.236478896335358e-06, + "loss": 2.3669, "step": 2457 }, { - "epoch": 0.7871897518014411, - "grad_norm": 0.5390625, - "learning_rate": 6.582502817869868e-06, - "loss": 2.2507, + "epoch": 0.9550035210412569, + "grad_norm": 0.609375, + "learning_rate": 1.2152952855190692e-06, + "loss": 2.311, "step": 2458 }, { - "epoch": 0.7875100080064051, - "grad_norm": 0.53125, - "learning_rate": 6.563607191074345e-06, - "loss": 2.264, + "epoch": 0.9553920497316724, + "grad_norm": 0.625, + "learning_rate": 1.1942935976126724e-06, + "loss": 2.4434, "step": 2459 }, { - "epoch": 0.7878302642113691, - "grad_norm": 0.408203125, - "learning_rate": 6.5447346249622926e-06, - "loss": 2.4022, + "epoch": 0.9557805784220879, + "grad_norm": 0.58984375, + "learning_rate": 1.1734738712930849e-06, + "loss": 2.3608, "step": 2460 }, { - "epoch": 0.788150520416333, - "grad_norm": 0.515625, - "learning_rate": 6.52588514313994e-06, - "loss": 2.6929, + "epoch": 0.9561691071125034, + "grad_norm": 0.625, + "learning_rate": 1.1528361449021475e-06, + "loss": 2.3522, "step": 2461 }, { - "epoch": 0.788470776621297, - "grad_norm": 0.5, - "learning_rate": 6.507058769184699e-06, - "loss": 2.1782, + "epoch": 0.9565576358029189, + "grad_norm": 0.5703125, + "learning_rate": 1.1323804564464802e-06, + "loss": 2.3984, "step": 2462 }, { - "epoch": 0.7887910328262611, - "grad_norm": 0.8359375, - "learning_rate": 6.488255526645029e-06, - "loss": 1.8834, + "epoch": 0.9569461644933343, + "grad_norm": 0.6328125, + "learning_rate": 1.1121068435974935e-06, + "loss": 2.3543, "step": 2463 }, { - "epoch": 0.789111289031225, - "grad_norm": 0.5234375, - "learning_rate": 6.46947543904049e-06, - "loss": 2.4088, + "epoch": 0.9573346931837498, + "grad_norm": 0.60546875, + "learning_rate": 1.0920153436912883e-06, + "loss": 2.2961, "step": 2464 }, { - "epoch": 0.789431545236189, - "grad_norm": 0.486328125, - "learning_rate": 6.450718529861663e-06, - "loss": 2.7434, + "epoch": 0.9577232218741653, + "grad_norm": 0.62109375, + "learning_rate": 1.072105993728556e-06, + "loss": 2.3472, "step": 2465 }, { - "epoch": 0.7897518014411529, - "grad_norm": 0.365234375, - "learning_rate": 6.431984822570147e-06, - "loss": 2.0478, + "epoch": 0.9581117505645808, + "grad_norm": 0.61328125, + "learning_rate": 1.0523788303745674e-06, + "loss": 2.3973, "step": 2466 }, { - "epoch": 0.7900720576461169, - "grad_norm": 0.9765625, - "learning_rate": 6.413274340598496e-06, - "loss": 2.5495, + "epoch": 0.9585002792549963, + "grad_norm": 0.60546875, + "learning_rate": 1.0328338899590616e-06, + "loss": 2.3392, "step": 2467 }, { - "epoch": 0.7903923138510809, - "grad_norm": 0.68359375, - "learning_rate": 6.394587107350258e-06, - "loss": 1.9307, + "epoch": 0.9588888079454118, + "grad_norm": 0.65234375, + "learning_rate": 1.0134712084762022e-06, + "loss": 2.3869, "step": 2468 }, { - "epoch": 0.7907125700560448, - "grad_norm": 0.37890625, - "learning_rate": 6.3759231461998525e-06, - "loss": 1.8534, + "epoch": 0.9592773366358271, + "grad_norm": 0.6171875, + "learning_rate": 9.94290821584498e-07, + "loss": 2.303, "step": 2469 }, { - "epoch": 0.7910328262610088, - "grad_norm": 0.357421875, - "learning_rate": 6.357282480492616e-06, - "loss": 1.9793, + "epoch": 0.9596658653262427, + "grad_norm": 0.58203125, + "learning_rate": 9.752927646067388e-07, + "loss": 2.3608, "step": 2470 }, { - "epoch": 0.7913530824659728, - "grad_norm": 0.5234375, - "learning_rate": 6.338665133544744e-06, - "loss": 2.3047, + "epoch": 0.9600543940166582, + "grad_norm": 0.65625, + "learning_rate": 9.564770725299376e-07, + "loss": 2.3699, "step": 2471 }, { - "epoch": 0.7916733386709367, - "grad_norm": 0.75, - "learning_rate": 6.32007112864326e-06, - "loss": 2.4341, + "epoch": 0.9604429227070737, + "grad_norm": 0.609375, + "learning_rate": 9.378437800052764e-07, + "loss": 2.3729, "step": 2472 }, { - "epoch": 0.7919935948759007, - "grad_norm": 0.453125, - "learning_rate": 6.3015004890459915e-06, - "loss": 1.6601, + "epoch": 0.9608314513974892, + "grad_norm": 0.58203125, + "learning_rate": 9.193929213480057e-07, + "loss": 2.3335, "step": 2473 }, { - "epoch": 0.7923138510808647, - "grad_norm": 0.455078125, - "learning_rate": 6.282953237981551e-06, - "loss": 2.1922, + "epoch": 0.9612199800879047, + "grad_norm": 0.6171875, + "learning_rate": 9.011245305374006e-07, + "loss": 2.3212, "step": 2474 }, { - "epoch": 0.7926341072858286, - "grad_norm": 0.41796875, - "learning_rate": 6.264429398649263e-06, - "loss": 1.6936, + "epoch": 0.96160850877832, + "grad_norm": 0.5859375, + "learning_rate": 8.83038641216738e-07, + "loss": 2.3239, "step": 2475 }, { - "epoch": 0.7929543634907926, - "grad_norm": 0.48828125, - "learning_rate": 6.245928994219222e-06, - "loss": 1.8489, + "epoch": 0.9619970374687355, + "grad_norm": 0.609375, + "learning_rate": 8.651352866931639e-07, + "loss": 2.2481, "step": 2476 }, { - "epoch": 0.7932746196957566, - "grad_norm": 1.03125, - "learning_rate": 6.227452047832155e-06, - "loss": 2.7892, + "epoch": 0.962385566159151, + "grad_norm": 0.6484375, + "learning_rate": 8.47414499937671e-07, + "loss": 2.3728, "step": 2477 }, { - "epoch": 0.7935948759007205, - "grad_norm": 0.478515625, - "learning_rate": 6.208998582599476e-06, - "loss": 2.4221, + "epoch": 0.9627740948495666, + "grad_norm": 0.60546875, + "learning_rate": 8.298763135850429e-07, + "loss": 2.3857, "step": 2478 }, { - "epoch": 0.7939151321056845, - "grad_norm": 0.50390625, - "learning_rate": 6.1905686216032264e-06, - "loss": 2.2199, + "epoch": 0.963162623539982, + "grad_norm": 0.59765625, + "learning_rate": 8.125207599337769e-07, + "loss": 2.293, "step": 2479 }, { - "epoch": 0.7942353883106485, - "grad_norm": 0.703125, - "learning_rate": 6.1721621878960526e-06, - "loss": 2.4266, + "epoch": 0.9635511522303976, + "grad_norm": 0.5859375, + "learning_rate": 7.953478709460394e-07, + "loss": 2.347, "step": 2480 }, { - "epoch": 0.7945556445156124, - "grad_norm": 0.43359375, - "learning_rate": 6.153779304501139e-06, - "loss": 1.9723, + "epoch": 0.963939680920813, + "grad_norm": 0.62890625, + "learning_rate": 7.783576782476099e-07, + "loss": 2.3957, "step": 2481 }, { - "epoch": 0.7948759007205765, - "grad_norm": 0.515625, - "learning_rate": 6.135419994412272e-06, - "loss": 2.4696, + "epoch": 0.9643282096112284, + "grad_norm": 0.625, + "learning_rate": 7.615502131277819e-07, + "loss": 2.3651, "step": 2482 }, { - "epoch": 0.7951961569255405, - "grad_norm": 0.48046875, - "learning_rate": 6.1170842805936905e-06, - "loss": 2.081, + "epoch": 0.964716738301644, + "grad_norm": 0.5859375, + "learning_rate": 7.449255065393624e-07, + "loss": 2.2837, "step": 2483 }, { - "epoch": 0.7955164131305044, - "grad_norm": 0.9609375, - "learning_rate": 6.098772185980159e-06, - "loss": 2.4716, + "epoch": 0.9651052669920595, + "grad_norm": 0.59765625, + "learning_rate": 7.284835890985608e-07, + "loss": 2.3985, "step": 2484 }, { - "epoch": 0.7958366693354684, - "grad_norm": 0.65234375, - "learning_rate": 6.080483733476883e-06, - "loss": 1.9128, + "epoch": 0.965493795682475, + "grad_norm": 0.61328125, + "learning_rate": 7.122244910850006e-07, + "loss": 2.3838, "step": 2485 }, { - "epoch": 0.7961569255404324, - "grad_norm": 0.5625, - "learning_rate": 6.062218945959497e-06, - "loss": 2.4175, + "epoch": 0.9658823243728905, + "grad_norm": 0.6484375, + "learning_rate": 6.961482424415855e-07, + "loss": 2.3677, "step": 2486 }, { - "epoch": 0.7964771817453963, - "grad_norm": 0.76171875, - "learning_rate": 6.04397784627404e-06, - "loss": 2.0246, + "epoch": 0.9662708530633058, + "grad_norm": 0.60546875, + "learning_rate": 6.802548727745106e-07, + "loss": 2.4086, "step": 2487 }, { - "epoch": 0.7967974379503603, - "grad_norm": 0.447265625, - "learning_rate": 6.025760457236926e-06, - "loss": 2.3537, + "epoch": 0.9666593817537213, + "grad_norm": 0.6015625, + "learning_rate": 6.645444113531519e-07, + "loss": 2.3239, "step": 2488 }, { - "epoch": 0.7971176941553243, - "grad_norm": 0.8515625, - "learning_rate": 6.007566801634876e-06, - "loss": 2.7204, + "epoch": 0.9670479104441368, + "grad_norm": 0.58984375, + "learning_rate": 6.490168871100766e-07, + "loss": 2.3253, "step": 2489 }, { - "epoch": 0.7974379503602882, - "grad_norm": 0.48828125, - "learning_rate": 5.989396902224986e-06, - "loss": 2.1869, + "epoch": 0.9674364391345524, + "grad_norm": 0.60546875, + "learning_rate": 6.336723286408996e-07, + "loss": 2.3663, "step": 2490 }, { - "epoch": 0.7977582065652522, - "grad_norm": 0.5859375, - "learning_rate": 5.97125078173458e-06, - "loss": 2.5756, + "epoch": 0.9678249678249679, + "grad_norm": 0.6171875, + "learning_rate": 6.185107642043275e-07, + "loss": 2.3496, "step": 2491 }, { - "epoch": 0.7980784627702162, - "grad_norm": 0.51171875, - "learning_rate": 5.953128462861268e-06, - "loss": 2.2436, + "epoch": 0.9682134965153834, + "grad_norm": 0.59765625, + "learning_rate": 6.035322217220584e-07, + "loss": 2.377, "step": 2492 }, { - "epoch": 0.7983987189751801, - "grad_norm": 0.3984375, - "learning_rate": 5.935029968272885e-06, - "loss": 1.6643, + "epoch": 0.9686020252057987, + "grad_norm": 0.62890625, + "learning_rate": 5.88736728778727e-07, + "loss": 2.3933, "step": 2493 }, { - "epoch": 0.7987189751801441, - "grad_norm": 0.478515625, - "learning_rate": 5.916955320607465e-06, - "loss": 2.1245, + "epoch": 0.9689905538962142, + "grad_norm": 0.625, + "learning_rate": 5.741243126218487e-07, + "loss": 2.3243, "step": 2494 }, { - "epoch": 0.7990392313851081, - "grad_norm": 0.419921875, - "learning_rate": 5.898904542473197e-06, - "loss": 2.1199, + "epoch": 0.9693790825866297, + "grad_norm": 0.60546875, + "learning_rate": 5.596950001618085e-07, + "loss": 2.3747, "step": 2495 }, { - "epoch": 0.799359487590072, - "grad_norm": 0.4765625, - "learning_rate": 5.8808776564484524e-06, - "loss": 2.4544, + "epoch": 0.9697676112770452, + "grad_norm": 0.61328125, + "learning_rate": 5.45448817971772e-07, + "loss": 2.3768, "step": 2496 }, { - "epoch": 0.799679743795036, - "grad_norm": 0.58203125, - "learning_rate": 5.862874685081673e-06, - "loss": 2.386, + "epoch": 0.9701561399674608, + "grad_norm": 0.5859375, + "learning_rate": 5.313857922876419e-07, + "loss": 2.4069, "step": 2497 }, { - "epoch": 0.8, - "grad_norm": 0.60546875, - "learning_rate": 5.844895650891416e-06, - "loss": 2.2664, + "epoch": 0.9705446686578763, + "grad_norm": 0.59765625, + "learning_rate": 5.17505949008057e-07, + "loss": 2.3018, "step": 2498 }, { - "epoch": 0.8003202562049639, - "grad_norm": 0.453125, - "learning_rate": 5.826940576366291e-06, - "loss": 2.2293, + "epoch": 0.9709331973482916, + "grad_norm": 0.59765625, + "learning_rate": 5.038093136942367e-07, + "loss": 2.3644, "step": 2499 }, { - "epoch": 0.800640512409928, - "grad_norm": 0.474609375, - "learning_rate": 5.809009483964936e-06, - "loss": 1.8299, + "epoch": 0.9713217260387071, + "grad_norm": 0.61328125, + "learning_rate": 4.902959115700712e-07, + "loss": 2.325, "step": 2500 }, { - "epoch": 0.800960768614892, - "grad_norm": 0.4453125, - "learning_rate": 5.791102396115991e-06, - "loss": 2.4019, + "epoch": 0.9717102547291226, + "grad_norm": 0.625, + "learning_rate": 4.769657675219752e-07, + "loss": 2.4057, "step": 2501 }, { - "epoch": 0.8012810248198559, - "grad_norm": 0.494140625, - "learning_rate": 5.773219335218083e-06, - "loss": 2.5534, + "epoch": 0.9720987834195381, + "grad_norm": 0.59765625, + "learning_rate": 4.6381890609886736e-07, + "loss": 2.3302, "step": 2502 }, { - "epoch": 0.8016012810248199, - "grad_norm": 0.478515625, - "learning_rate": 5.75536032363975e-06, - "loss": 2.1044, + "epoch": 0.9724873121099536, + "grad_norm": 0.6015625, + "learning_rate": 4.508553515121472e-07, + "loss": 2.367, "step": 2503 }, { - "epoch": 0.8019215372297839, - "grad_norm": 0.359375, - "learning_rate": 5.737525383719499e-06, - "loss": 1.7572, + "epoch": 0.9728758408003692, + "grad_norm": 0.62890625, + "learning_rate": 4.380751276356176e-07, + "loss": 2.4321, "step": 2504 }, { - "epoch": 0.8022417934347478, - "grad_norm": 0.46484375, - "learning_rate": 5.719714537765688e-06, - "loss": 2.5762, + "epoch": 0.9732643694907845, + "grad_norm": 0.61328125, + "learning_rate": 4.25478258005485e-07, + "loss": 2.4378, "step": 2505 }, { - "epoch": 0.8025620496397118, - "grad_norm": 0.81640625, - "learning_rate": 5.7019278080565535e-06, - "loss": 2.3808, + "epoch": 0.9736528981812, + "grad_norm": 0.6015625, + "learning_rate": 4.130647658202591e-07, + "loss": 2.2761, "step": 2506 }, { - "epoch": 0.8028823058446758, - "grad_norm": 0.4765625, - "learning_rate": 5.684165216840165e-06, - "loss": 2.4309, + "epoch": 0.9740414268716155, + "grad_norm": 0.640625, + "learning_rate": 4.008346739407642e-07, + "loss": 2.3617, "step": 2507 }, { - "epoch": 0.8032025620496397, - "grad_norm": 2.203125, - "learning_rate": 5.666426786334405e-06, - "loss": 2.4435, + "epoch": 0.974429955562031, + "grad_norm": 0.65625, + "learning_rate": 3.887880048900394e-07, + "loss": 2.3557, "step": 2508 }, { - "epoch": 0.8035228182546037, - "grad_norm": 0.48046875, - "learning_rate": 5.648712538726911e-06, - "loss": 2.1253, + "epoch": 0.9748184842524465, + "grad_norm": 0.5859375, + "learning_rate": 3.7692478085337155e-07, + "loss": 2.3727, "step": 2509 }, { - "epoch": 0.8038430744595677, - "grad_norm": 0.62109375, - "learning_rate": 5.631022496175115e-06, - "loss": 2.5646, + "epoch": 0.975207012942862, + "grad_norm": 0.58984375, + "learning_rate": 3.652450236781957e-07, + "loss": 2.3027, "step": 2510 }, { - "epoch": 0.8041633306645316, - "grad_norm": 0.490234375, - "learning_rate": 5.61335668080612e-06, - "loss": 2.106, + "epoch": 0.9755955416332774, + "grad_norm": 0.61328125, + "learning_rate": 3.5374875487405037e-07, + "loss": 2.3671, "step": 2511 }, { - "epoch": 0.8044835868694956, - "grad_norm": 0.5078125, - "learning_rate": 5.595715114716782e-06, - "loss": 2.2236, + "epoch": 0.9759840703236929, + "grad_norm": 0.61328125, + "learning_rate": 3.424359956126e-07, + "loss": 2.3165, "step": 2512 }, { - "epoch": 0.8048038430744595, - "grad_norm": 0.53125, - "learning_rate": 5.578097819973577e-06, - "loss": 2.451, + "epoch": 0.9763725990141084, + "grad_norm": 0.625, + "learning_rate": 3.313067667275238e-07, + "loss": 2.3788, "step": 2513 }, { - "epoch": 0.8051240992794235, - "grad_norm": 0.84765625, - "learning_rate": 5.560504818612641e-06, - "loss": 2.1605, + "epoch": 0.976761127704524, + "grad_norm": 0.63671875, + "learning_rate": 3.2036108871452695e-07, + "loss": 2.3492, "step": 2514 }, { - "epoch": 0.8054443554843875, - "grad_norm": 0.4765625, - "learning_rate": 5.5429361326397285e-06, - "loss": 2.5384, + "epoch": 0.9771496563949394, + "grad_norm": 0.57421875, + "learning_rate": 3.0959898173128497e-07, + "loss": 2.2973, "step": 2515 }, { - "epoch": 0.8057646116893514, - "grad_norm": 0.36328125, - "learning_rate": 5.525391784030182e-06, - "loss": 2.0079, + "epoch": 0.977538185085355, + "grad_norm": 0.61328125, + "learning_rate": 2.990204655973994e-07, + "loss": 2.4172, "step": 2516 }, { - "epoch": 0.8060848678943154, - "grad_norm": 0.375, - "learning_rate": 5.507871794728872e-06, - "loss": 1.5529, + "epoch": 0.9779267137757703, + "grad_norm": 0.62890625, + "learning_rate": 2.8862555979437544e-07, + "loss": 2.3124, "step": 2517 }, { - "epoch": 0.8064051240992794, - "grad_norm": 0.388671875, - "learning_rate": 5.4903761866502475e-06, - "loss": 2.1881, + "epoch": 0.9783152424661858, + "grad_norm": 0.5859375, + "learning_rate": 2.784142834655667e-07, + "loss": 2.3208, "step": 2518 }, { - "epoch": 0.8067253803042433, - "grad_norm": 0.6015625, - "learning_rate": 5.47290498167822e-06, - "loss": 2.1702, + "epoch": 0.9787037711566013, + "grad_norm": 0.609375, + "learning_rate": 2.6838665541616404e-07, + "loss": 2.3884, "step": 2519 }, { - "epoch": 0.8070456365092074, - "grad_norm": 0.40234375, - "learning_rate": 5.455458201666197e-06, - "loss": 2.2788, + "epoch": 0.9790922998470168, + "grad_norm": 0.59375, + "learning_rate": 2.585426941131619e-07, + "loss": 2.4062, "step": 2520 }, { - "epoch": 0.8073658927141714, - "grad_norm": 0.57421875, - "learning_rate": 5.43803586843703e-06, - "loss": 2.2958, + "epoch": 0.9794808285374323, + "grad_norm": 0.64453125, + "learning_rate": 2.488824176852922e-07, + "loss": 2.3114, "step": 2521 }, { - "epoch": 0.8076861489191353, - "grad_norm": 0.57421875, - "learning_rate": 5.420638003783002e-06, - "loss": 2.4007, + "epoch": 0.9798693572278478, + "grad_norm": 0.62109375, + "learning_rate": 2.394058439230129e-07, + "loss": 2.3523, "step": 2522 }, { - "epoch": 0.8080064051240993, - "grad_norm": 0.416015625, - "learning_rate": 5.403264629465757e-06, - "loss": 2.1257, + "epoch": 0.9802578859182632, + "grad_norm": 0.609375, + "learning_rate": 2.3011299027847487e-07, + "loss": 2.3356, "step": 2523 }, { - "epoch": 0.8083266613290633, - "grad_norm": 0.40234375, - "learning_rate": 5.3859157672163564e-06, - "loss": 2.2356, + "epoch": 0.9806464146086787, + "grad_norm": 0.59765625, + "learning_rate": 2.210038738654996e-07, + "loss": 2.3744, "step": 2524 }, { - "epoch": 0.8086469175340272, - "grad_norm": 0.58984375, - "learning_rate": 5.368591438735149e-06, - "loss": 2.5254, + "epoch": 0.9810349432990942, + "grad_norm": 0.59375, + "learning_rate": 2.1207851145953473e-07, + "loss": 2.2672, "step": 2525 }, { - "epoch": 0.8089671737389912, - "grad_norm": 0.5, - "learning_rate": 5.3512916656918485e-06, - "loss": 2.1586, + "epoch": 0.9814234719895097, + "grad_norm": 0.59375, + "learning_rate": 2.033369194975987e-07, + "loss": 2.3416, "step": 2526 }, { - "epoch": 0.8092874299439552, - "grad_norm": 1.1875, - "learning_rate": 5.334016469725409e-06, - "loss": 2.3058, + "epoch": 0.9818120006799252, + "grad_norm": 0.62109375, + "learning_rate": 1.9477911407831396e-07, + "loss": 2.3808, "step": 2527 }, { - "epoch": 0.8096076861489191, - "grad_norm": 0.99609375, - "learning_rate": 5.31676587244406e-06, - "loss": 2.061, + "epoch": 0.9822005293703406, + "grad_norm": 0.58984375, + "learning_rate": 1.86405110961807e-07, + "loss": 2.3057, "step": 2528 }, { - "epoch": 0.8099279423538831, - "grad_norm": 0.470703125, - "learning_rate": 5.299539895425271e-06, - "loss": 2.6985, + "epoch": 0.9825890580607561, + "grad_norm": 0.609375, + "learning_rate": 1.7821492556973075e-07, + "loss": 2.4372, "step": 2529 }, { - "epoch": 0.8102481985588471, - "grad_norm": 0.5234375, - "learning_rate": 5.282338560215711e-06, - "loss": 1.9756, + "epoch": 0.9829775867511716, + "grad_norm": 0.609375, + "learning_rate": 1.7020857298521986e-07, + "loss": 2.3674, "step": 2530 }, { - "epoch": 0.810568454763811, - "grad_norm": 0.4375, - "learning_rate": 5.265161888331205e-06, - "loss": 2.1155, + "epoch": 0.9833661154415871, + "grad_norm": 0.640625, + "learning_rate": 1.6238606795284662e-07, + "loss": 2.3223, "step": 2531 }, { - "epoch": 0.810888710968775, - "grad_norm": 1.1953125, - "learning_rate": 5.248009901256776e-06, - "loss": 2.2378, + "epoch": 0.9837546441320026, + "grad_norm": 0.59375, + "learning_rate": 1.547474248786096e-07, + "loss": 2.2912, "step": 2532 }, { - "epoch": 0.811208967173739, - "grad_norm": 0.55078125, - "learning_rate": 5.230882620446517e-06, - "loss": 2.1956, + "epoch": 0.9841431728224181, + "grad_norm": 0.625, + "learning_rate": 1.4729265782993383e-07, + "loss": 2.3836, "step": 2533 }, { - "epoch": 0.8115292233787029, - "grad_norm": 0.416015625, - "learning_rate": 5.213780067323654e-06, - "loss": 2.4922, + "epoch": 0.9845317015128335, + "grad_norm": 0.59765625, + "learning_rate": 1.400217805355708e-07, + "loss": 2.3461, "step": 2534 }, { - "epoch": 0.8118494795836669, - "grad_norm": 0.85546875, - "learning_rate": 5.19670226328047e-06, - "loss": 2.0935, + "epoch": 0.984920230203249, + "grad_norm": 0.6875, + "learning_rate": 1.3293480638565392e-07, + "loss": 2.3769, "step": 2535 }, { - "epoch": 0.8121697357886309, - "grad_norm": 0.470703125, - "learning_rate": 5.1796492296782905e-06, - "loss": 1.7568, + "epoch": 0.9853087588936645, + "grad_norm": 0.6015625, + "learning_rate": 1.2603174843162092e-07, + "loss": 2.2917, "step": 2536 }, { - "epoch": 0.8124899919935948, - "grad_norm": 0.498046875, - "learning_rate": 5.162620987847466e-06, - "loss": 2.527, + "epoch": 0.98569728758408, + "grad_norm": 0.61328125, + "learning_rate": 1.1931261938621375e-07, + "loss": 2.3128, "step": 2537 }, { - "epoch": 0.8128102481985588, - "grad_norm": 0.5234375, - "learning_rate": 5.145617559087332e-06, - "loss": 2.5302, + "epoch": 0.9860858162744955, + "grad_norm": 0.6171875, + "learning_rate": 1.1277743162345644e-07, + "loss": 2.344, "step": 2538 }, { - "epoch": 0.8131305044035229, - "grad_norm": 0.56640625, - "learning_rate": 5.128638964666166e-06, - "loss": 2.2782, + "epoch": 0.986474344964911, + "grad_norm": 0.65234375, + "learning_rate": 1.0642619717859959e-07, + "loss": 2.328, "step": 2539 }, { - "epoch": 0.8134507606084868, - "grad_norm": 0.56640625, - "learning_rate": 5.111685225821233e-06, - "loss": 2.4849, + "epoch": 0.9868628736553264, + "grad_norm": 0.59765625, + "learning_rate": 1.002589277481536e-07, + "loss": 2.3472, "step": 2540 }, { - "epoch": 0.8137710168134508, - "grad_norm": 0.41796875, - "learning_rate": 5.0947563637586545e-06, - "loss": 2.1828, + "epoch": 0.9872514023457419, + "grad_norm": 0.6484375, + "learning_rate": 9.427563468982215e-08, + "loss": 2.3188, "step": 2541 }, { - "epoch": 0.8140912730184148, - "grad_norm": 0.453125, - "learning_rate": 5.077852399653463e-06, - "loss": 2.3564, + "epoch": 0.9876399310361574, + "grad_norm": 0.62109375, + "learning_rate": 8.847632902250214e-08, + "loss": 2.3605, "step": 2542 }, { - "epoch": 0.8144115292233787, - "grad_norm": 0.48828125, - "learning_rate": 5.060973354649548e-06, - "loss": 1.8303, + "epoch": 0.9880284597265729, + "grad_norm": 0.64453125, + "learning_rate": 8.28610214262393e-08, + "loss": 2.3247, "step": 2543 }, { - "epoch": 0.8147317854283427, - "grad_norm": 0.59375, - "learning_rate": 5.044119249859627e-06, - "loss": 2.3438, + "epoch": 0.9884169884169884, + "grad_norm": 0.60546875, + "learning_rate": 7.742972224225043e-08, + "loss": 2.3225, "step": 2544 }, { - "epoch": 0.8150520416333067, - "grad_norm": 0.4375, - "learning_rate": 5.027290106365204e-06, - "loss": 1.9135, + "epoch": 0.9888055171074039, + "grad_norm": 0.640625, + "learning_rate": 7.21824414728789e-08, + "loss": 2.3744, "step": 2545 }, { - "epoch": 0.8153722978382706, - "grad_norm": 0.3515625, - "learning_rate": 5.010485945216601e-06, - "loss": 2.1691, + "epoch": 0.9891940457978193, + "grad_norm": 0.61328125, + "learning_rate": 6.711918878155033e-08, + "loss": 2.2812, "step": 2546 }, { - "epoch": 0.8156925540432346, - "grad_norm": 0.86328125, - "learning_rate": 4.993706787432844e-06, - "loss": 2.5787, + "epoch": 0.9895825744882348, + "grad_norm": 0.62109375, + "learning_rate": 6.223997349282807e-08, + "loss": 2.3866, "step": 2547 }, { - "epoch": 0.8160128102481986, - "grad_norm": 0.482421875, - "learning_rate": 4.976952654001718e-06, - "loss": 2.1237, + "epoch": 0.9899711031786503, + "grad_norm": 0.65234375, + "learning_rate": 5.754480459229106e-08, + "loss": 2.377, "step": 2548 }, { - "epoch": 0.8163330664531625, - "grad_norm": 0.44921875, - "learning_rate": 4.960223565879696e-06, - "loss": 2.3624, + "epoch": 0.9903596318690658, + "grad_norm": 0.62890625, + "learning_rate": 5.303369072664488e-08, + "loss": 2.3846, "step": 2549 }, { - "epoch": 0.8166533226581265, - "grad_norm": 0.53125, - "learning_rate": 4.94351954399192e-06, - "loss": 1.776, + "epoch": 0.9907481605594813, + "grad_norm": 6.3125, + "learning_rate": 4.8706640203577406e-08, + "loss": 2.2776, "step": 2550 }, { - "epoch": 0.8169735788630905, - "grad_norm": 0.53125, - "learning_rate": 4.926840609232183e-06, - "loss": 2.3213, + "epoch": 0.9911366892498968, + "grad_norm": 0.60546875, + "learning_rate": 4.456366099183651e-08, + "loss": 2.3427, "step": 2551 }, { - "epoch": 0.8172938350680544, - "grad_norm": 1.0234375, - "learning_rate": 4.910186782462908e-06, - "loss": 2.7406, + "epoch": 0.9915252179403122, + "grad_norm": 0.59765625, + "learning_rate": 4.0604760721174586e-08, + "loss": 2.3697, "step": 2552 }, { - "epoch": 0.8176140912730184, - "grad_norm": 0.431640625, - "learning_rate": 4.8935580845150765e-06, - "loss": 2.2117, + "epoch": 0.9919137466307277, + "grad_norm": 0.609375, + "learning_rate": 3.6829946682348516e-08, + "loss": 2.3838, "step": 2553 }, { - "epoch": 0.8179343474779824, - "grad_norm": 0.38671875, - "learning_rate": 4.876954536188297e-06, - "loss": 1.6675, + "epoch": 0.9923022753211432, + "grad_norm": 0.58984375, + "learning_rate": 3.3239225827086386e-08, + "loss": 2.2705, "step": 2554 }, { - "epoch": 0.8182546036829463, - "grad_norm": 0.67578125, - "learning_rate": 4.860376158250659e-06, - "loss": 2.0331, + "epoch": 0.9926908040115587, + "grad_norm": 0.61328125, + "learning_rate": 2.983260476808747e-08, + "loss": 2.3065, "step": 2555 }, { - "epoch": 0.8185748598879103, - "grad_norm": 0.4375, - "learning_rate": 4.84382297143881e-06, - "loss": 2.0956, + "epoch": 0.9930793327019742, + "grad_norm": 0.63671875, + "learning_rate": 2.6610089779044445e-08, + "loss": 2.3795, "step": 2556 }, { - "epoch": 0.8188951160928744, - "grad_norm": 0.59375, - "learning_rate": 4.827294996457873e-06, - "loss": 2.4513, + "epoch": 0.9934678613923897, + "grad_norm": 0.61328125, + "learning_rate": 2.3571686794543467e-08, + "loss": 2.3717, "step": 2557 }, { - "epoch": 0.8192153722978383, - "grad_norm": 1.46875, - "learning_rate": 4.810792253981439e-06, - "loss": 2.3092, + "epoch": 0.9938563900828051, + "grad_norm": 0.62890625, + "learning_rate": 2.0717401410164096e-08, + "loss": 2.3575, "step": 2558 }, { - "epoch": 0.8195356285028023, - "grad_norm": 0.34765625, - "learning_rate": 4.79431476465152e-06, - "loss": 1.9795, + "epoch": 0.9942449187732206, + "grad_norm": 0.578125, + "learning_rate": 1.8047238882379357e-08, + "loss": 2.3786, "step": 2559 }, { - "epoch": 0.8198558847077662, - "grad_norm": 0.43359375, - "learning_rate": 4.777862549078579e-06, - "loss": 2.2858, + "epoch": 0.9946334474636361, + "grad_norm": 0.61328125, + "learning_rate": 1.556120412857798e-08, + "loss": 2.4276, "step": 2560 }, { - "epoch": 0.8201761409127302, - "grad_norm": 0.421875, - "learning_rate": 4.7614356278414235e-06, - "loss": 2.7803, + "epoch": 0.9950219761540516, + "grad_norm": 0.66015625, + "learning_rate": 1.3259301727075458e-08, + "loss": 2.3399, "step": 2561 }, { - "epoch": 0.8204963971176942, - "grad_norm": 1.0625, - "learning_rate": 4.745034021487252e-06, - "loss": 2.0644, + "epoch": 0.9954105048444671, + "grad_norm": 0.609375, + "learning_rate": 1.1141535917069678e-08, + "loss": 2.3158, "step": 2562 }, { - "epoch": 0.8208166533226581, - "grad_norm": 1.234375, - "learning_rate": 4.728657750531581e-06, - "loss": 1.9754, + "epoch": 0.9957990335348826, + "grad_norm": 0.61328125, + "learning_rate": 9.207910598674208e-09, + "loss": 2.317, "step": 2563 }, { - "epoch": 0.8211369095276221, - "grad_norm": 1.390625, - "learning_rate": 4.712306835458249e-06, - "loss": 2.4305, + "epoch": 0.996187562225298, + "grad_norm": 0.6171875, + "learning_rate": 7.458429332862781e-09, + "loss": 2.3655, "step": 2564 }, { - "epoch": 0.8214571657325861, - "grad_norm": 0.83203125, - "learning_rate": 4.695981296719376e-06, - "loss": 2.0684, + "epoch": 0.9965760909157135, + "grad_norm": 0.66015625, + "learning_rate": 5.893095341502619e-09, + "loss": 2.3816, "step": 2565 }, { - "epoch": 0.82177742193755, - "grad_norm": 0.40234375, - "learning_rate": 4.679681154735341e-06, - "loss": 2.2563, + "epoch": 0.996964619606129, + "grad_norm": 0.66796875, + "learning_rate": 4.511911507321109e-09, + "loss": 2.2994, "step": 2566 }, { - "epoch": 0.822097678142514, - "grad_norm": 0.5078125, - "learning_rate": 4.6634064298947315e-06, - "loss": 2.162, + "epoch": 0.9973531482965445, + "grad_norm": 0.59375, + "learning_rate": 3.31488037392802e-09, + "loss": 2.3347, "step": 2567 }, { - "epoch": 0.822417934347478, - "grad_norm": 0.44140625, - "learning_rate": 4.647157142554396e-06, - "loss": 2.3221, + "epoch": 0.99774167698696, + "grad_norm": 0.58984375, + "learning_rate": 2.3020041457821885e-09, + "loss": 2.2885, "step": 2568 }, { - "epoch": 0.8227381905524419, - "grad_norm": 0.453125, - "learning_rate": 4.6309333130393125e-06, - "loss": 2.307, + "epoch": 0.9981302056773755, + "grad_norm": 0.671875, + "learning_rate": 1.4732846882137274e-09, + "loss": 2.3607, "step": 2569 }, { - "epoch": 0.8230584467574059, - "grad_norm": 0.5703125, - "learning_rate": 4.614734961642644e-06, - "loss": 2.0686, + "epoch": 0.9985187343677909, + "grad_norm": 0.6328125, + "learning_rate": 8.287235273907179e-10, + "loss": 2.3445, "step": 2570 }, { - "epoch": 0.8233787029623699, - "grad_norm": 0.6328125, - "learning_rate": 4.598562108625676e-06, - "loss": 2.1714, + "epoch": 0.9989072630582064, + "grad_norm": 0.62890625, + "learning_rate": 3.683218503636177e-10, + "loss": 2.3514, "step": 2571 }, { - "epoch": 0.8236989591673338, - "grad_norm": 1.6796875, - "learning_rate": 4.582414774217811e-06, - "loss": 2.3787, + "epoch": 0.9992957917486219, + "grad_norm": 0.58984375, + "learning_rate": 9.208050498754617e-11, + "loss": 2.3164, "step": 2572 }, { - "epoch": 0.8240192153722978, - "grad_norm": 0.4609375, - "learning_rate": 4.5662929786165014e-06, - "loss": 2.2512, + "epoch": 0.9996843204390374, + "grad_norm": 0.6171875, + "learning_rate": 0.0, + "loss": 2.3925, "step": 2573 }, { - "epoch": 0.8243394715772618, - "grad_norm": 0.376953125, - "learning_rate": 4.550196741987306e-06, - "loss": 2.0636, - "step": 2574 - }, - { - "epoch": 0.8246597277822257, - "grad_norm": 0.50390625, - "learning_rate": 4.534126084463755e-06, - "loss": 1.9996, - "step": 2575 - }, - { - "epoch": 0.8249799839871897, - "grad_norm": 0.45703125, - "learning_rate": 4.518081026147439e-06, - "loss": 1.8507, - "step": 2576 - }, - { - "epoch": 0.8253002401921538, - "grad_norm": 0.58984375, - "learning_rate": 4.502061587107889e-06, - "loss": 2.1717, - "step": 2577 - }, - { - "epoch": 0.8256204963971177, - "grad_norm": 2.109375, - "learning_rate": 4.4860677873826025e-06, - "loss": 2.3655, - "step": 2578 - }, - { - "epoch": 0.8259407526020817, - "grad_norm": 0.443359375, - "learning_rate": 4.4700996469770165e-06, - "loss": 2.014, - "step": 2579 - }, - { - "epoch": 0.8262610088070457, - "grad_norm": 0.515625, - "learning_rate": 4.454157185864469e-06, - "loss": 2.4824, - "step": 2580 - }, - { - "epoch": 0.8265812650120096, - "grad_norm": 0.578125, - "learning_rate": 4.438240423986154e-06, - "loss": 2.2044, - "step": 2581 - }, - { - "epoch": 0.8269015212169736, - "grad_norm": 0.455078125, - "learning_rate": 4.422349381251162e-06, - "loss": 2.2686, - "step": 2582 - }, - { - "epoch": 0.8272217774219376, - "grad_norm": 0.400390625, - "learning_rate": 4.406484077536377e-06, - "loss": 2.5848, - "step": 2583 - }, - { - "epoch": 0.8275420336269015, - "grad_norm": 0.56640625, - "learning_rate": 4.3906445326865085e-06, - "loss": 1.95, - "step": 2584 - }, - { - "epoch": 0.8278622898318655, - "grad_norm": 0.50390625, - "learning_rate": 4.374830766514037e-06, - "loss": 1.8195, - "step": 2585 - }, - { - "epoch": 0.8281825460368295, - "grad_norm": 0.453125, - "learning_rate": 4.3590427987992126e-06, - "loss": 1.9135, - "step": 2586 - }, - { - "epoch": 0.8285028022417934, - "grad_norm": 0.474609375, - "learning_rate": 4.3432806492899826e-06, - "loss": 2.2409, - "step": 2587 - }, - { - "epoch": 0.8288230584467574, - "grad_norm": 0.466796875, - "learning_rate": 4.327544337702047e-06, - "loss": 1.7218, - "step": 2588 - }, - { - "epoch": 0.8291433146517214, - "grad_norm": 0.5078125, - "learning_rate": 4.3118338837187415e-06, - "loss": 2.4292, - "step": 2589 - }, - { - "epoch": 0.8294635708566853, - "grad_norm": 0.490234375, - "learning_rate": 4.296149306991098e-06, - "loss": 2.1419, - "step": 2590 - }, - { - "epoch": 0.8297838270616493, - "grad_norm": 0.640625, - "learning_rate": 4.280490627137751e-06, - "loss": 2.5364, - "step": 2591 - }, - { - "epoch": 0.8301040832666133, - "grad_norm": 0.4296875, - "learning_rate": 4.264857863744956e-06, - "loss": 2.121, - "step": 2592 - }, - { - "epoch": 0.8304243394715772, - "grad_norm": 1.4140625, - "learning_rate": 4.24925103636655e-06, - "loss": 2.5428, - "step": 2593 - }, - { - "epoch": 0.8307445956765412, - "grad_norm": 0.5703125, - "learning_rate": 4.233670164523937e-06, - "loss": 2.0423, - "step": 2594 - }, - { - "epoch": 0.8310648518815053, - "grad_norm": 0.419921875, - "learning_rate": 4.218115267706021e-06, - "loss": 2.4526, - "step": 2595 - }, - { - "epoch": 0.8313851080864691, - "grad_norm": 0.396484375, - "learning_rate": 4.2025863653692706e-06, - "loss": 2.3808, - "step": 2596 - }, - { - "epoch": 0.8317053642914332, - "grad_norm": 0.53515625, - "learning_rate": 4.1870834769375925e-06, - "loss": 2.3936, - "step": 2597 - }, - { - "epoch": 0.8320256204963972, - "grad_norm": 0.52734375, - "learning_rate": 4.171606621802377e-06, - "loss": 2.443, - "step": 2598 - }, - { - "epoch": 0.8323458767013611, - "grad_norm": 1.296875, - "learning_rate": 4.1561558193224435e-06, - "loss": 2.0169, - "step": 2599 - }, - { - "epoch": 0.8326661329063251, - "grad_norm": 0.37890625, - "learning_rate": 4.14073108882404e-06, - "loss": 2.1607, - "step": 2600 - }, - { - "epoch": 0.8329863891112891, - "grad_norm": 0.5234375, - "learning_rate": 4.125332449600766e-06, - "loss": 2.3544, - "step": 2601 - }, - { - "epoch": 0.833306645316253, - "grad_norm": 0.52734375, - "learning_rate": 4.109959920913637e-06, - "loss": 2.0641, - "step": 2602 - }, - { - "epoch": 0.833626901521217, - "grad_norm": 0.373046875, - "learning_rate": 4.094613521990956e-06, - "loss": 2.1128, - "step": 2603 - }, - { - "epoch": 0.833947157726181, - "grad_norm": 0.52734375, - "learning_rate": 4.079293272028395e-06, - "loss": 1.7678, - "step": 2604 - }, - { - "epoch": 0.8342674139311449, - "grad_norm": 0.76953125, - "learning_rate": 4.063999190188869e-06, - "loss": 2.2154, - "step": 2605 - }, - { - "epoch": 0.8345876701361089, - "grad_norm": 0.482421875, - "learning_rate": 4.04873129560259e-06, - "loss": 2.4884, - "step": 2606 - }, - { - "epoch": 0.8349079263410728, - "grad_norm": 0.5546875, - "learning_rate": 4.033489607367011e-06, - "loss": 2.6735, - "step": 2607 - }, - { - "epoch": 0.8352281825460368, - "grad_norm": 0.69140625, - "learning_rate": 4.018274144546802e-06, - "loss": 2.579, - "step": 2608 - }, - { - "epoch": 0.8355484387510008, - "grad_norm": 0.609375, - "learning_rate": 4.003084926173819e-06, - "loss": 2.5984, - "step": 2609 - }, - { - "epoch": 0.8358686949559647, - "grad_norm": 0.55078125, - "learning_rate": 3.987921971247122e-06, - "loss": 2.7333, - "step": 2610 - }, - { - "epoch": 0.8361889511609287, - "grad_norm": 0.439453125, - "learning_rate": 3.972785298732881e-06, - "loss": 2.0657, - "step": 2611 - }, - { - "epoch": 0.8365092073658927, - "grad_norm": 0.546875, - "learning_rate": 3.9576749275644175e-06, - "loss": 2.1598, - "step": 2612 - }, - { - "epoch": 0.8368294635708566, - "grad_norm": 0.45703125, - "learning_rate": 3.942590876642146e-06, - "loss": 2.1292, - "step": 2613 - }, - { - "epoch": 0.8371497197758206, - "grad_norm": 1.203125, - "learning_rate": 3.927533164833561e-06, - "loss": 2.3755, - "step": 2614 - }, - { - "epoch": 0.8374699759807847, - "grad_norm": 0.359375, - "learning_rate": 3.91250181097321e-06, - "loss": 2.1559, - "step": 2615 - }, - { - "epoch": 0.8377902321857486, - "grad_norm": 0.462890625, - "learning_rate": 3.897496833862677e-06, - "loss": 2.4086, - "step": 2616 - }, - { - "epoch": 0.8381104883907126, - "grad_norm": 2.890625, - "learning_rate": 3.882518252270531e-06, - "loss": 2.1332, - "step": 2617 - }, - { - "epoch": 0.8384307445956766, - "grad_norm": 0.5546875, - "learning_rate": 3.867566084932367e-06, - "loss": 1.9732, - "step": 2618 - }, - { - "epoch": 0.8387510008006405, - "grad_norm": 0.87890625, - "learning_rate": 3.852640350550696e-06, - "loss": 2.4946, - "step": 2619 - }, - { - "epoch": 0.8390712570056045, - "grad_norm": 0.5546875, - "learning_rate": 3.837741067794992e-06, - "loss": 2.6402, - "step": 2620 - }, - { - "epoch": 0.8393915132105685, - "grad_norm": 1.625, - "learning_rate": 3.822868255301637e-06, - "loss": 2.6491, - "step": 2621 - }, - { - "epoch": 0.8397117694155324, - "grad_norm": 0.64453125, - "learning_rate": 3.808021931673905e-06, - "loss": 2.4719, - "step": 2622 - }, - { - "epoch": 0.8400320256204964, - "grad_norm": 0.474609375, - "learning_rate": 3.7932021154819225e-06, - "loss": 1.9543, - "step": 2623 - }, - { - "epoch": 0.8403522818254604, - "grad_norm": 0.4453125, - "learning_rate": 3.7784088252626914e-06, - "loss": 1.8275, - "step": 2624 - }, - { - "epoch": 0.8406725380304243, - "grad_norm": 0.796875, - "learning_rate": 3.7636420795199996e-06, - "loss": 1.8942, - "step": 2625 - }, - { - "epoch": 0.8409927942353883, - "grad_norm": 0.48828125, - "learning_rate": 3.7489018967244494e-06, - "loss": 1.9354, - "step": 2626 - }, - { - "epoch": 0.8413130504403523, - "grad_norm": 0.3515625, - "learning_rate": 3.7341882953134206e-06, - "loss": 1.8982, - "step": 2627 - }, - { - "epoch": 0.8416333066453162, - "grad_norm": 1.0625, - "learning_rate": 3.7195012936910383e-06, - "loss": 2.8696, - "step": 2628 - }, - { - "epoch": 0.8419535628502802, - "grad_norm": 0.6796875, - "learning_rate": 3.7048409102281577e-06, - "loss": 2.3575, - "step": 2629 - }, - { - "epoch": 0.8422738190552442, - "grad_norm": 0.82421875, - "learning_rate": 3.690207163262341e-06, - "loss": 2.4408, - "step": 2630 - }, - { - "epoch": 0.8425940752602081, - "grad_norm": 0.41015625, - "learning_rate": 3.675600071097818e-06, - "loss": 2.1077, - "step": 2631 - }, - { - "epoch": 0.8429143314651721, - "grad_norm": 0.45703125, - "learning_rate": 3.6610196520055124e-06, - "loss": 2.1918, - "step": 2632 - }, - { - "epoch": 0.8432345876701361, - "grad_norm": 0.57421875, - "learning_rate": 3.646465924222947e-06, - "loss": 2.4826, - "step": 2633 - }, - { - "epoch": 0.8435548438751, - "grad_norm": 0.494140625, - "learning_rate": 3.631938905954277e-06, - "loss": 2.3311, - "step": 2634 - }, - { - "epoch": 0.8438751000800641, - "grad_norm": 0.447265625, - "learning_rate": 3.6174386153702473e-06, - "loss": 2.4787, - "step": 2635 - }, - { - "epoch": 0.8441953562850281, - "grad_norm": 0.44140625, - "learning_rate": 3.6029650706081775e-06, - "loss": 1.9944, - "step": 2636 - }, - { - "epoch": 0.844515612489992, - "grad_norm": 0.478515625, - "learning_rate": 3.588518289771908e-06, - "loss": 2.4228, - "step": 2637 - }, - { - "epoch": 0.844835868694956, - "grad_norm": 0.42578125, - "learning_rate": 3.5740982909318406e-06, - "loss": 2.4106, - "step": 2638 - }, - { - "epoch": 0.84515612489992, - "grad_norm": 0.515625, - "learning_rate": 3.5597050921248347e-06, - "loss": 2.4529, - "step": 2639 - }, - { - "epoch": 0.8454763811048839, - "grad_norm": 0.478515625, - "learning_rate": 3.5453387113542737e-06, - "loss": 2.5682, - "step": 2640 - }, - { - "epoch": 0.8457966373098479, - "grad_norm": 1.609375, - "learning_rate": 3.5309991665899545e-06, - "loss": 2.0301, - "step": 2641 - }, - { - "epoch": 0.8461168935148119, - "grad_norm": 0.42578125, - "learning_rate": 3.516686475768127e-06, - "loss": 2.1756, - "step": 2642 - }, - { - "epoch": 0.8464371497197758, - "grad_norm": 0.8046875, - "learning_rate": 3.502400656791455e-06, - "loss": 1.9134, - "step": 2643 - }, - { - "epoch": 0.8467574059247398, - "grad_norm": 0.890625, - "learning_rate": 3.4881417275289847e-06, - "loss": 2.3704, - "step": 2644 - }, - { - "epoch": 0.8470776621297038, - "grad_norm": 0.57421875, - "learning_rate": 3.4739097058161114e-06, - "loss": 2.5802, - "step": 2645 - }, - { - "epoch": 0.8473979183346677, - "grad_norm": 0.45703125, - "learning_rate": 3.459704609454614e-06, - "loss": 1.5905, - "step": 2646 - }, - { - "epoch": 0.8477181745396317, - "grad_norm": 0.55859375, - "learning_rate": 3.445526456212553e-06, - "loss": 2.2604, - "step": 2647 - }, - { - "epoch": 0.8480384307445957, - "grad_norm": 0.482421875, - "learning_rate": 3.4313752638243075e-06, - "loss": 2.3032, - "step": 2648 - }, - { - "epoch": 0.8483586869495596, - "grad_norm": 0.447265625, - "learning_rate": 3.41725104999053e-06, - "loss": 2.215, - "step": 2649 - }, - { - "epoch": 0.8486789431545236, - "grad_norm": 0.56640625, - "learning_rate": 3.4031538323781254e-06, - "loss": 2.3237, - "step": 2650 - }, - { - "epoch": 0.8489991993594876, - "grad_norm": 0.6328125, - "learning_rate": 3.389083628620235e-06, - "loss": 1.5601, - "step": 2651 - }, - { - "epoch": 0.8493194555644515, - "grad_norm": 0.283203125, - "learning_rate": 3.3750404563162085e-06, - "loss": 2.3487, - "step": 2652 - }, - { - "epoch": 0.8496397117694156, - "grad_norm": 0.55859375, - "learning_rate": 3.3610243330315733e-06, - "loss": 1.9077, - "step": 2653 - }, - { - "epoch": 0.8499599679743794, - "grad_norm": 0.427734375, - "learning_rate": 3.347035276298052e-06, - "loss": 1.879, - "step": 2654 - }, - { - "epoch": 0.8502802241793435, - "grad_norm": 0.546875, - "learning_rate": 3.3330733036134765e-06, - "loss": 2.0285, - "step": 2655 - }, - { - "epoch": 0.8506004803843075, - "grad_norm": 1.21875, - "learning_rate": 3.3191384324418236e-06, - "loss": 2.5377, - "step": 2656 - }, - { - "epoch": 0.8509207365892714, - "grad_norm": 0.498046875, - "learning_rate": 3.3052306802131644e-06, - "loss": 1.8052, - "step": 2657 - }, - { - "epoch": 0.8512409927942354, - "grad_norm": 0.5390625, - "learning_rate": 3.291350064323659e-06, - "loss": 1.8514, - "step": 2658 - }, - { - "epoch": 0.8515612489991994, - "grad_norm": 0.490234375, - "learning_rate": 3.277496602135496e-06, - "loss": 2.2399, - "step": 2659 - }, - { - "epoch": 0.8518815052041633, - "grad_norm": 0.419921875, - "learning_rate": 3.2636703109769406e-06, - "loss": 2.274, - "step": 2660 - }, - { - "epoch": 0.8522017614091273, - "grad_norm": 0.578125, - "learning_rate": 3.249871208142238e-06, - "loss": 2.2057, - "step": 2661 - }, - { - "epoch": 0.8525220176140913, - "grad_norm": 0.55078125, - "learning_rate": 3.236099310891638e-06, - "loss": 2.5783, - "step": 2662 - }, - { - "epoch": 0.8528422738190552, - "grad_norm": 0.65625, - "learning_rate": 3.2223546364513636e-06, - "loss": 2.2508, - "step": 2663 - }, - { - "epoch": 0.8531625300240192, - "grad_norm": 0.46875, - "learning_rate": 3.2086372020135815e-06, - "loss": 2.5538, - "step": 2664 - }, - { - "epoch": 0.8534827862289832, - "grad_norm": 0.625, - "learning_rate": 3.194947024736392e-06, - "loss": 2.0804, - "step": 2665 - }, - { - "epoch": 0.8538030424339471, - "grad_norm": 0.640625, - "learning_rate": 3.1812841217438023e-06, - "loss": 2.5366, - "step": 2666 - }, - { - "epoch": 0.8541232986389111, - "grad_norm": 0.66015625, - "learning_rate": 3.1676485101256844e-06, - "loss": 2.1118, - "step": 2667 - }, - { - "epoch": 0.8544435548438751, - "grad_norm": 0.578125, - "learning_rate": 3.1540402069378084e-06, - "loss": 2.3653, - "step": 2668 - }, - { - "epoch": 0.854763811048839, - "grad_norm": 0.5625, - "learning_rate": 3.140459229201753e-06, - "loss": 2.4334, - "step": 2669 - }, - { - "epoch": 0.855084067253803, - "grad_norm": 0.5, - "learning_rate": 3.1269055939049356e-06, - "loss": 2.5745, - "step": 2670 - }, - { - "epoch": 0.855404323458767, - "grad_norm": 0.376953125, - "learning_rate": 3.11337931800057e-06, - "loss": 1.7917, - "step": 2671 - }, - { - "epoch": 0.8557245796637309, - "grad_norm": 0.61328125, - "learning_rate": 3.0998804184076495e-06, - "loss": 2.3825, - "step": 2672 - }, - { - "epoch": 0.856044835868695, - "grad_norm": 0.59765625, - "learning_rate": 3.0864089120109186e-06, - "loss": 2.1879, - "step": 2673 - }, - { - "epoch": 0.856365092073659, - "grad_norm": 0.408203125, - "learning_rate": 3.07296481566087e-06, - "loss": 2.2279, - "step": 2674 - }, - { - "epoch": 0.8566853482786229, - "grad_norm": 0.62890625, - "learning_rate": 3.059548146173691e-06, - "loss": 2.1929, - "step": 2675 - }, - { - "epoch": 0.8570056044835869, - "grad_norm": 0.625, - "learning_rate": 3.046158920331277e-06, - "loss": 2.2886, - "step": 2676 - }, - { - "epoch": 0.8573258606885509, - "grad_norm": 0.55078125, - "learning_rate": 3.032797154881198e-06, - "loss": 2.8405, - "step": 2677 - }, - { - "epoch": 0.8576461168935148, - "grad_norm": 0.48046875, - "learning_rate": 3.019462866536668e-06, - "loss": 1.712, - "step": 2678 - }, - { - "epoch": 0.8579663730984788, - "grad_norm": 0.47265625, - "learning_rate": 3.0061560719765376e-06, - "loss": 2.4549, - "step": 2679 - }, - { - "epoch": 0.8582866293034428, - "grad_norm": 0.61328125, - "learning_rate": 2.9928767878452698e-06, - "loss": 2.2048, - "step": 2680 - }, - { - "epoch": 0.8586068855084067, - "grad_norm": 0.828125, - "learning_rate": 2.9796250307528973e-06, - "loss": 2.1138, - "step": 2681 - }, - { - "epoch": 0.8589271417133707, - "grad_norm": 3.390625, - "learning_rate": 2.9664008172750557e-06, - "loss": 2.0606, - "step": 2682 - }, - { - "epoch": 0.8592473979183347, - "grad_norm": 0.5859375, - "learning_rate": 2.9532041639528967e-06, - "loss": 2.1709, - "step": 2683 - }, - { - "epoch": 0.8595676541232986, - "grad_norm": 0.61328125, - "learning_rate": 2.9400350872931138e-06, - "loss": 2.5132, - "step": 2684 - }, - { - "epoch": 0.8598879103282626, - "grad_norm": 0.5625, - "learning_rate": 2.926893603767908e-06, - "loss": 2.2391, - "step": 2685 - }, - { - "epoch": 0.8602081665332266, - "grad_norm": 0.62890625, - "learning_rate": 2.9137797298149584e-06, - "loss": 2.4567, - "step": 2686 - }, - { - "epoch": 0.8605284227381905, - "grad_norm": 0.61328125, - "learning_rate": 2.9006934818374187e-06, - "loss": 2.511, - "step": 2687 - }, - { - "epoch": 0.8608486789431545, - "grad_norm": 0.48046875, - "learning_rate": 2.8876348762038846e-06, - "loss": 2.4784, - "step": 2688 - }, - { - "epoch": 0.8611689351481185, - "grad_norm": 0.625, - "learning_rate": 2.874603929248365e-06, - "loss": 1.7693, - "step": 2689 - }, - { - "epoch": 0.8614891913530824, - "grad_norm": 0.421875, - "learning_rate": 2.8616006572702857e-06, - "loss": 2.4669, - "step": 2690 - }, - { - "epoch": 0.8618094475580464, - "grad_norm": 0.453125, - "learning_rate": 2.848625076534453e-06, - "loss": 1.9782, - "step": 2691 - }, - { - "epoch": 0.8621297037630105, - "grad_norm": 0.51953125, - "learning_rate": 2.835677203271037e-06, - "loss": 2.2473, - "step": 2692 - }, - { - "epoch": 0.8624499599679744, - "grad_norm": 0.6328125, - "learning_rate": 2.8227570536755425e-06, - "loss": 2.3449, - "step": 2693 - }, - { - "epoch": 0.8627702161729384, - "grad_norm": 0.50390625, - "learning_rate": 2.8098646439088106e-06, - "loss": 2.292, - "step": 2694 - }, - { - "epoch": 0.8630904723779024, - "grad_norm": 0.498046875, - "learning_rate": 2.796999990096974e-06, - "loss": 2.395, - "step": 2695 - }, - { - "epoch": 0.8634107285828663, - "grad_norm": 0.43359375, - "learning_rate": 2.7841631083314584e-06, - "loss": 2.19, - "step": 2696 - }, - { - "epoch": 0.8637309847878303, - "grad_norm": 0.357421875, - "learning_rate": 2.771354014668931e-06, - "loss": 1.5913, - "step": 2697 - }, - { - "epoch": 0.8640512409927943, - "grad_norm": 0.470703125, - "learning_rate": 2.75857272513132e-06, - "loss": 2.1674, - "step": 2698 - }, - { - "epoch": 0.8643714971977582, - "grad_norm": 0.77734375, - "learning_rate": 2.74581925570577e-06, - "loss": 2.3245, - "step": 2699 - }, - { - "epoch": 0.8646917534027222, - "grad_norm": 0.51953125, - "learning_rate": 2.733093622344629e-06, - "loss": 1.8666, - "step": 2700 - }, - { - "epoch": 0.8650120096076861, - "grad_norm": 0.51171875, - "learning_rate": 2.7203958409654227e-06, - "loss": 2.2844, - "step": 2701 - }, - { - "epoch": 0.8653322658126501, - "grad_norm": 0.94140625, - "learning_rate": 2.707725927450844e-06, - "loss": 1.9818, - "step": 2702 - }, - { - "epoch": 0.8656525220176141, - "grad_norm": 0.578125, - "learning_rate": 2.6950838976487287e-06, - "loss": 1.9651, - "step": 2703 - }, - { - "epoch": 0.865972778222578, - "grad_norm": 1.3203125, - "learning_rate": 2.68246976737202e-06, - "loss": 2.1429, - "step": 2704 - }, - { - "epoch": 0.866293034427542, - "grad_norm": 0.7109375, - "learning_rate": 2.6698835523987832e-06, - "loss": 2.1287, - "step": 2705 - }, - { - "epoch": 0.866613290632506, - "grad_norm": 0.5234375, - "learning_rate": 2.657325268472155e-06, - "loss": 2.2263, - "step": 2706 - }, - { - "epoch": 0.8669335468374699, - "grad_norm": 0.51953125, - "learning_rate": 2.6447949313003456e-06, - "loss": 2.1328, - "step": 2707 - }, - { - "epoch": 0.8672538030424339, - "grad_norm": 0.640625, - "learning_rate": 2.6322925565565954e-06, - "loss": 2.1588, - "step": 2708 - }, - { - "epoch": 0.8675740592473979, - "grad_norm": 0.55078125, - "learning_rate": 2.619818159879178e-06, - "loss": 2.1747, - "step": 2709 - }, - { - "epoch": 0.8678943154523618, - "grad_norm": 0.54296875, - "learning_rate": 2.607371756871374e-06, - "loss": 1.9288, - "step": 2710 - }, - { - "epoch": 0.8682145716573259, - "grad_norm": 0.59375, - "learning_rate": 2.594953363101435e-06, - "loss": 2.3231, - "step": 2711 - }, - { - "epoch": 0.8685348278622899, - "grad_norm": 0.55859375, - "learning_rate": 2.5825629941025887e-06, - "loss": 2.2347, - "step": 2712 - }, - { - "epoch": 0.8688550840672538, - "grad_norm": 0.50390625, - "learning_rate": 2.5702006653730072e-06, - "loss": 2.5465, - "step": 2713 - }, - { - "epoch": 0.8691753402722178, - "grad_norm": 0.984375, - "learning_rate": 2.557866392375788e-06, - "loss": 2.163, - "step": 2714 - }, - { - "epoch": 0.8694955964771818, - "grad_norm": 0.56640625, - "learning_rate": 2.545560190538937e-06, - "loss": 2.5304, - "step": 2715 - }, - { - "epoch": 0.8698158526821457, - "grad_norm": 0.6015625, - "learning_rate": 2.5332820752553498e-06, - "loss": 2.4132, - "step": 2716 - }, - { - "epoch": 0.8701361088871097, - "grad_norm": 0.419921875, - "learning_rate": 2.521032061882786e-06, - "loss": 2.2646, - "step": 2717 - }, - { - "epoch": 0.8704563650920737, - "grad_norm": 0.83984375, - "learning_rate": 2.5088101657438605e-06, - "loss": 2.1072, - "step": 2718 - }, - { - "epoch": 0.8707766212970376, - "grad_norm": 0.42578125, - "learning_rate": 2.4966164021260092e-06, - "loss": 2.3744, - "step": 2719 - }, - { - "epoch": 0.8710968775020016, - "grad_norm": 0.53515625, - "learning_rate": 2.4844507862814835e-06, - "loss": 1.145, - "step": 2720 - }, - { - "epoch": 0.8714171337069656, - "grad_norm": 0.396484375, - "learning_rate": 2.472313333427334e-06, - "loss": 1.9044, - "step": 2721 - }, - { - "epoch": 0.8717373899119295, - "grad_norm": 0.515625, - "learning_rate": 2.4602040587453774e-06, - "loss": 1.9285, - "step": 2722 - }, - { - "epoch": 0.8720576461168935, - "grad_norm": 0.87109375, - "learning_rate": 2.4481229773821835e-06, - "loss": 2.2034, - "step": 2723 - }, - { - "epoch": 0.8723779023218575, - "grad_norm": 0.6171875, - "learning_rate": 2.436070104449059e-06, - "loss": 2.475, - "step": 2724 - }, - { - "epoch": 0.8726981585268214, - "grad_norm": 0.671875, - "learning_rate": 2.4240454550220367e-06, - "loss": 2.528, - "step": 2725 - }, - { - "epoch": 0.8730184147317854, - "grad_norm": 0.55859375, - "learning_rate": 2.412049044141823e-06, - "loss": 2.067, - "step": 2726 - }, - { - "epoch": 0.8733386709367494, - "grad_norm": 0.51171875, - "learning_rate": 2.4000808868138212e-06, - "loss": 1.8833, - "step": 2727 - }, - { - "epoch": 0.8736589271417133, - "grad_norm": 0.59765625, - "learning_rate": 2.3881409980080954e-06, - "loss": 2.4556, - "step": 2728 - }, - { - "epoch": 0.8739791833466773, - "grad_norm": 0.50390625, - "learning_rate": 2.3762293926593433e-06, - "loss": 1.9098, - "step": 2729 - }, - { - "epoch": 0.8742994395516414, - "grad_norm": 0.5859375, - "learning_rate": 2.3643460856668823e-06, - "loss": 2.4019, - "step": 2730 - }, - { - "epoch": 0.8746196957566053, - "grad_norm": 0.5078125, - "learning_rate": 2.3524910918946484e-06, - "loss": 2.2306, - "step": 2731 - }, - { - "epoch": 0.8749399519615693, - "grad_norm": 0.484375, - "learning_rate": 2.340664426171149e-06, - "loss": 2.2344, - "step": 2732 - }, - { - "epoch": 0.8752602081665333, - "grad_norm": 0.453125, - "learning_rate": 2.3288661032894564e-06, - "loss": 2.1087, - "step": 2733 - }, - { - "epoch": 0.8755804643714972, - "grad_norm": 0.345703125, - "learning_rate": 2.3170961380071993e-06, - "loss": 2.0231, - "step": 2734 - }, - { - "epoch": 0.8759007205764612, - "grad_norm": 0.640625, - "learning_rate": 2.305354545046537e-06, - "loss": 2.5689, - "step": 2735 - }, - { - "epoch": 0.8762209767814252, - "grad_norm": 0.609375, - "learning_rate": 2.2936413390941325e-06, - "loss": 2.1854, - "step": 2736 - }, - { - "epoch": 0.8765412329863891, - "grad_norm": 0.49609375, - "learning_rate": 2.2819565348011466e-06, - "loss": 2.2984, - "step": 2737 - }, - { - "epoch": 0.8768614891913531, - "grad_norm": 0.5234375, - "learning_rate": 2.2703001467832137e-06, - "loss": 2.3775, - "step": 2738 - }, - { - "epoch": 0.8771817453963171, - "grad_norm": 0.69140625, - "learning_rate": 2.258672189620431e-06, - "loss": 2.5423, - "step": 2739 - }, - { - "epoch": 0.877502001601281, - "grad_norm": 1.5, - "learning_rate": 2.2470726778573146e-06, - "loss": 1.8138, - "step": 2740 - }, - { - "epoch": 0.877822257806245, - "grad_norm": 0.4375, - "learning_rate": 2.235501626002817e-06, - "loss": 1.9844, - "step": 2741 - }, - { - "epoch": 0.878142514011209, - "grad_norm": 0.5703125, - "learning_rate": 2.2239590485302887e-06, - "loss": 1.8902, - "step": 2742 - }, - { - "epoch": 0.8784627702161729, - "grad_norm": 0.59765625, - "learning_rate": 2.212444959877466e-06, - "loss": 1.7303, - "step": 2743 - }, - { - "epoch": 0.8787830264211369, - "grad_norm": 0.53125, - "learning_rate": 2.2009593744464456e-06, - "loss": 2.1603, - "step": 2744 - }, - { - "epoch": 0.8791032826261009, - "grad_norm": 0.494140625, - "learning_rate": 2.1895023066036745e-06, - "loss": 1.5858, - "step": 2745 - }, - { - "epoch": 0.8794235388310648, - "grad_norm": 0.6640625, - "learning_rate": 2.1780737706799255e-06, - "loss": 2.3251, - "step": 2746 - }, - { - "epoch": 0.8797437950360288, - "grad_norm": 0.6171875, - "learning_rate": 2.1666737809702963e-06, - "loss": 2.1973, - "step": 2747 - }, - { - "epoch": 0.8800640512409927, - "grad_norm": 0.5546875, - "learning_rate": 2.1553023517341566e-06, - "loss": 2.2133, - "step": 2748 - }, - { - "epoch": 0.8803843074459567, - "grad_norm": 0.44921875, - "learning_rate": 2.1439594971951666e-06, - "loss": 2.7235, - "step": 2749 - }, - { - "epoch": 0.8807045636509208, - "grad_norm": 0.470703125, - "learning_rate": 2.1326452315412413e-06, - "loss": 1.9469, - "step": 2750 - }, - { - "epoch": 0.8810248198558847, - "grad_norm": 2.046875, - "learning_rate": 2.1213595689245386e-06, - "loss": 2.0951, - "step": 2751 - }, - { - "epoch": 0.8813450760608487, - "grad_norm": 1.6796875, - "learning_rate": 2.110102523461435e-06, - "loss": 2.1507, - "step": 2752 - }, - { - "epoch": 0.8816653322658127, - "grad_norm": 0.494140625, - "learning_rate": 2.098874109232521e-06, - "loss": 2.0099, - "step": 2753 - }, - { - "epoch": 0.8819855884707766, - "grad_norm": 0.98828125, - "learning_rate": 2.08767434028255e-06, - "loss": 1.7282, - "step": 2754 - }, - { - "epoch": 0.8823058446757406, - "grad_norm": 6.3125, - "learning_rate": 2.07650323062048e-06, - "loss": 2.3165, - "step": 2755 - }, - { - "epoch": 0.8826261008807046, - "grad_norm": 0.419921875, - "learning_rate": 2.065360794219395e-06, - "loss": 2.093, - "step": 2756 - }, - { - "epoch": 0.8829463570856685, - "grad_norm": 0.890625, - "learning_rate": 2.054247045016522e-06, - "loss": 1.8592, - "step": 2757 - }, - { - "epoch": 0.8832666132906325, - "grad_norm": 0.7265625, - "learning_rate": 2.0431619969132065e-06, - "loss": 2.0496, - "step": 2758 - }, - { - "epoch": 0.8835868694955965, - "grad_norm": 0.6328125, - "learning_rate": 2.032105663774897e-06, - "loss": 2.394, - "step": 2759 - }, - { - "epoch": 0.8839071257005604, - "grad_norm": 0.3671875, - "learning_rate": 2.021078059431114e-06, - "loss": 1.8352, - "step": 2760 - }, - { - "epoch": 0.8842273819055244, - "grad_norm": 0.96484375, - "learning_rate": 2.010079197675463e-06, - "loss": 1.8796, - "step": 2761 - }, - { - "epoch": 0.8845476381104884, - "grad_norm": 0.71875, - "learning_rate": 1.9991090922655694e-06, - "loss": 1.7734, - "step": 2762 - }, - { - "epoch": 0.8848678943154523, - "grad_norm": 0.99609375, - "learning_rate": 1.9881677569231094e-06, - "loss": 2.5961, - "step": 2763 - }, - { - "epoch": 0.8851881505204163, - "grad_norm": 0.48046875, - "learning_rate": 1.977255205333775e-06, - "loss": 1.9614, - "step": 2764 - }, - { - "epoch": 0.8855084067253803, - "grad_norm": 0.60546875, - "learning_rate": 1.966371451147239e-06, - "loss": 2.529, - "step": 2765 - }, - { - "epoch": 0.8858286629303442, - "grad_norm": 0.640625, - "learning_rate": 1.9555165079771703e-06, - "loss": 2.4828, - "step": 2766 - }, - { - "epoch": 0.8861489191353082, - "grad_norm": 0.75390625, - "learning_rate": 1.9446903894011937e-06, - "loss": 2.0968, - "step": 2767 - }, - { - "epoch": 0.8864691753402723, - "grad_norm": 0.47265625, - "learning_rate": 1.933893108960866e-06, - "loss": 1.7931, - "step": 2768 - }, - { - "epoch": 0.8867894315452362, - "grad_norm": 0.6484375, - "learning_rate": 1.9231246801617033e-06, - "loss": 2.7387, - "step": 2769 - }, - { - "epoch": 0.8871096877502002, - "grad_norm": 0.486328125, - "learning_rate": 1.9123851164731004e-06, - "loss": 2.1077, - "step": 2770 - }, - { - "epoch": 0.8874299439551642, - "grad_norm": 0.458984375, - "learning_rate": 1.901674431328368e-06, - "loss": 2.2398, - "step": 2771 - }, - { - "epoch": 0.8877502001601281, - "grad_norm": 0.458984375, - "learning_rate": 1.8909926381246862e-06, - "loss": 2.5055, - "step": 2772 - }, - { - "epoch": 0.8880704563650921, - "grad_norm": 0.41015625, - "learning_rate": 1.8803397502231034e-06, - "loss": 2.0305, - "step": 2773 - }, - { - "epoch": 0.8883907125700561, - "grad_norm": 0.47265625, - "learning_rate": 1.8697157809485055e-06, - "loss": 2.3018, - "step": 2774 - }, - { - "epoch": 0.88871096877502, - "grad_norm": 0.56640625, - "learning_rate": 1.8591207435896096e-06, - "loss": 2.3575, - "step": 2775 - }, - { - "epoch": 0.889031224979984, - "grad_norm": 0.486328125, - "learning_rate": 1.8485546513989372e-06, - "loss": 2.1141, - "step": 2776 - }, - { - "epoch": 0.889351481184948, - "grad_norm": 0.6640625, - "learning_rate": 1.8380175175928216e-06, - "loss": 1.9234, - "step": 2777 - }, - { - "epoch": 0.8896717373899119, - "grad_norm": 0.455078125, - "learning_rate": 1.8275093553513533e-06, - "loss": 2.3495, - "step": 2778 - }, - { - "epoch": 0.8899919935948759, - "grad_norm": 0.4921875, - "learning_rate": 1.817030177818399e-06, - "loss": 2.2586, - "step": 2779 - }, - { - "epoch": 0.8903122497998399, - "grad_norm": 0.50390625, - "learning_rate": 1.8065799981015625e-06, - "loss": 2.1218, - "step": 2780 - }, - { - "epoch": 0.8906325060048038, - "grad_norm": 0.5234375, - "learning_rate": 1.7961588292721848e-06, - "loss": 2.3084, - "step": 2781 - }, - { - "epoch": 0.8909527622097678, - "grad_norm": 0.59375, - "learning_rate": 1.7857666843653088e-06, - "loss": 1.9461, - "step": 2782 - }, - { - "epoch": 0.8912730184147318, - "grad_norm": 0.490234375, - "learning_rate": 1.775403576379689e-06, - "loss": 2.3538, - "step": 2783 - }, - { - "epoch": 0.8915932746196957, - "grad_norm": 0.470703125, - "learning_rate": 1.76506951827774e-06, - "loss": 2.2233, - "step": 2784 - }, - { - "epoch": 0.8919135308246597, - "grad_norm": 0.435546875, - "learning_rate": 1.7547645229855497e-06, - "loss": 2.5818, - "step": 2785 - }, - { - "epoch": 0.8922337870296237, - "grad_norm": 0.68359375, - "learning_rate": 1.7444886033928603e-06, - "loss": 2.1235, - "step": 2786 - }, - { - "epoch": 0.8925540432345876, - "grad_norm": 0.44140625, - "learning_rate": 1.7342417723530374e-06, - "loss": 2.3278, - "step": 2787 - }, - { - "epoch": 0.8928742994395517, - "grad_norm": 0.439453125, - "learning_rate": 1.7240240426830617e-06, - "loss": 2.1486, - "step": 2788 - }, - { - "epoch": 0.8931945556445157, - "grad_norm": 1.515625, - "learning_rate": 1.713835427163521e-06, - "loss": 2.7315, - "step": 2789 - }, - { - "epoch": 0.8935148118494796, - "grad_norm": 0.408203125, - "learning_rate": 1.7036759385385682e-06, - "loss": 1.8048, - "step": 2790 - }, - { - "epoch": 0.8938350680544436, - "grad_norm": 0.890625, - "learning_rate": 1.693545589515952e-06, - "loss": 2.6576, - "step": 2791 - }, - { - "epoch": 0.8941553242594076, - "grad_norm": 0.412109375, - "learning_rate": 1.6834443927669446e-06, - "loss": 1.56, - "step": 2792 - }, - { - "epoch": 0.8944755804643715, - "grad_norm": 0.4296875, - "learning_rate": 1.6733723609263674e-06, - "loss": 2.1219, - "step": 2793 - }, - { - "epoch": 0.8947958366693355, - "grad_norm": 0.83203125, - "learning_rate": 1.6633295065925647e-06, - "loss": 1.3158, - "step": 2794 - }, - { - "epoch": 0.8951160928742994, - "grad_norm": 0.494140625, - "learning_rate": 1.653315842327377e-06, - "loss": 2.4756, - "step": 2795 - }, - { - "epoch": 0.8954363490792634, - "grad_norm": 0.3828125, - "learning_rate": 1.6433313806561379e-06, - "loss": 2.2544, - "step": 2796 - }, - { - "epoch": 0.8957566052842274, - "grad_norm": 0.419921875, - "learning_rate": 1.6333761340676518e-06, - "loss": 2.0936, - "step": 2797 - }, - { - "epoch": 0.8960768614891913, - "grad_norm": 0.423828125, - "learning_rate": 1.6234501150141745e-06, - "loss": 2.3256, - "step": 2798 - }, - { - "epoch": 0.8963971176941553, - "grad_norm": 0.546875, - "learning_rate": 1.6135533359114247e-06, - "loss": 2.2462, - "step": 2799 - }, - { - "epoch": 0.8967173738991193, - "grad_norm": 0.4609375, - "learning_rate": 1.6036858091385137e-06, - "loss": 2.377, - "step": 2800 - }, - { - "epoch": 0.8970376301040832, - "grad_norm": 0.70703125, - "learning_rate": 1.593847547037991e-06, - "loss": 2.2872, - "step": 2801 - }, - { - "epoch": 0.8973578863090472, - "grad_norm": 0.5, - "learning_rate": 1.5840385619157904e-06, - "loss": 1.8975, - "step": 2802 - }, - { - "epoch": 0.8976781425140112, - "grad_norm": 0.5625, - "learning_rate": 1.574258866041231e-06, - "loss": 2.7625, - "step": 2803 - }, - { - "epoch": 0.8979983987189751, - "grad_norm": 0.498046875, - "learning_rate": 1.5645084716469777e-06, - "loss": 2.2334, - "step": 2804 - }, - { - "epoch": 0.8983186549239391, - "grad_norm": 2.90625, - "learning_rate": 1.5547873909290716e-06, - "loss": 2.3039, - "step": 2805 - }, - { - "epoch": 0.8986389111289032, - "grad_norm": 0.44140625, - "learning_rate": 1.5450956360468698e-06, - "loss": 1.9124, - "step": 2806 - }, - { - "epoch": 0.898959167333867, - "grad_norm": 0.455078125, - "learning_rate": 1.5354332191230502e-06, - "loss": 1.9301, - "step": 2807 - }, - { - "epoch": 0.8992794235388311, - "grad_norm": 0.498046875, - "learning_rate": 1.5258001522435978e-06, - "loss": 2.5829, - "step": 2808 - }, - { - "epoch": 0.8995996797437951, - "grad_norm": 0.5390625, - "learning_rate": 1.516196447457785e-06, - "loss": 2.4198, - "step": 2809 - }, - { - "epoch": 0.899919935948759, - "grad_norm": 0.7578125, - "learning_rate": 1.5066221167781557e-06, - "loss": 2.4602, - "step": 2810 - }, - { - "epoch": 0.900240192153723, - "grad_norm": 0.515625, - "learning_rate": 1.4970771721805165e-06, - "loss": 2.4408, - "step": 2811 - }, - { - "epoch": 0.900560448358687, - "grad_norm": 0.55859375, - "learning_rate": 1.4875616256039031e-06, - "loss": 2.3519, - "step": 2812 - }, - { - "epoch": 0.9008807045636509, - "grad_norm": 0.31640625, - "learning_rate": 1.478075488950606e-06, - "loss": 1.9272, - "step": 2813 - }, - { - "epoch": 0.9012009607686149, - "grad_norm": 0.42578125, - "learning_rate": 1.468618774086103e-06, - "loss": 2.1546, - "step": 2814 - }, - { - "epoch": 0.9015212169735789, - "grad_norm": 0.5390625, - "learning_rate": 1.4591914928390793e-06, - "loss": 2.2757, - "step": 2815 - }, - { - "epoch": 0.9018414731785428, - "grad_norm": 0.5234375, - "learning_rate": 1.449793657001408e-06, - "loss": 2.3683, - "step": 2816 - }, - { - "epoch": 0.9021617293835068, - "grad_norm": 0.50390625, - "learning_rate": 1.4404252783281308e-06, - "loss": 1.7778, - "step": 2817 - }, - { - "epoch": 0.9024819855884708, - "grad_norm": 0.326171875, - "learning_rate": 1.431086368537432e-06, - "loss": 2.2228, - "step": 2818 - }, - { - "epoch": 0.9028022417934347, - "grad_norm": 0.515625, - "learning_rate": 1.4217769393106567e-06, - "loss": 2.0747, - "step": 2819 - }, - { - "epoch": 0.9031224979983987, - "grad_norm": 0.4453125, - "learning_rate": 1.4124970022922513e-06, - "loss": 1.7582, - "step": 2820 - }, - { - "epoch": 0.9034427542033627, - "grad_norm": 0.6484375, - "learning_rate": 1.4032465690897944e-06, - "loss": 2.561, - "step": 2821 - }, - { - "epoch": 0.9037630104083266, - "grad_norm": 0.515625, - "learning_rate": 1.3940256512739447e-06, - "loss": 1.9619, - "step": 2822 - }, - { - "epoch": 0.9040832666132906, - "grad_norm": 0.478515625, - "learning_rate": 1.3848342603784453e-06, - "loss": 2.1804, - "step": 2823 - }, - { - "epoch": 0.9044035228182546, - "grad_norm": 0.5546875, - "learning_rate": 1.3756724079001165e-06, - "loss": 2.6895, - "step": 2824 - }, - { - "epoch": 0.9047237790232185, - "grad_norm": 0.421875, - "learning_rate": 1.366540105298819e-06, - "loss": 1.9138, - "step": 2825 - }, - { - "epoch": 0.9050440352281826, - "grad_norm": 0.396484375, - "learning_rate": 1.3574373639974518e-06, - "loss": 2.2131, - "step": 2826 - }, - { - "epoch": 0.9053642914331466, - "grad_norm": 0.412109375, - "learning_rate": 1.3483641953819542e-06, - "loss": 2.1942, - "step": 2827 - }, - { - "epoch": 0.9056845476381105, - "grad_norm": 0.75, - "learning_rate": 1.3393206108012562e-06, - "loss": 1.9268, - "step": 2828 - }, - { - "epoch": 0.9060048038430745, - "grad_norm": 0.435546875, - "learning_rate": 1.33030662156729e-06, - "loss": 2.5872, - "step": 2829 - }, - { - "epoch": 0.9063250600480385, - "grad_norm": 0.5703125, - "learning_rate": 1.3213222389549724e-06, - "loss": 2.0611, - "step": 2830 - }, - { - "epoch": 0.9066453162530024, - "grad_norm": 0.58203125, - "learning_rate": 1.3123674742021862e-06, - "loss": 2.4414, - "step": 2831 - }, - { - "epoch": 0.9069655724579664, - "grad_norm": 0.61328125, - "learning_rate": 1.303442338509761e-06, - "loss": 2.0825, - "step": 2832 - }, - { - "epoch": 0.9072858286629304, - "grad_norm": 2.046875, - "learning_rate": 1.2945468430414803e-06, - "loss": 1.8121, - "step": 2833 - }, - { - "epoch": 0.9076060848678943, - "grad_norm": 0.51953125, - "learning_rate": 1.2856809989240326e-06, - "loss": 2.3212, - "step": 2834 - }, - { - "epoch": 0.9079263410728583, - "grad_norm": 0.484375, - "learning_rate": 1.2768448172470444e-06, - "loss": 2.4806, - "step": 2835 - }, - { - "epoch": 0.9082465972778223, - "grad_norm": 0.5625, - "learning_rate": 1.2680383090630077e-06, - "loss": 2.4582, - "step": 2836 - }, - { - "epoch": 0.9085668534827862, - "grad_norm": 0.55859375, - "learning_rate": 1.2592614853873224e-06, - "loss": 2.4116, - "step": 2837 - }, - { - "epoch": 0.9088871096877502, - "grad_norm": 0.49609375, - "learning_rate": 1.2505143571982515e-06, - "loss": 2.1586, - "step": 2838 - }, - { - "epoch": 0.9092073658927142, - "grad_norm": 0.640625, - "learning_rate": 1.2417969354369147e-06, - "loss": 2.3296, - "step": 2839 - }, - { - "epoch": 0.9095276220976781, - "grad_norm": 0.69140625, - "learning_rate": 1.2331092310072623e-06, - "loss": 1.7858, - "step": 2840 - }, - { - "epoch": 0.9098478783026421, - "grad_norm": 0.470703125, - "learning_rate": 1.2244512547761016e-06, - "loss": 2.1908, - "step": 2841 - }, - { - "epoch": 0.910168134507606, - "grad_norm": 0.46484375, - "learning_rate": 1.2158230175730201e-06, - "loss": 2.1441, - "step": 2842 - }, - { - "epoch": 0.91048839071257, - "grad_norm": 0.64453125, - "learning_rate": 1.2072245301904345e-06, - "loss": 2.2306, - "step": 2843 - }, - { - "epoch": 0.910808646917534, - "grad_norm": 0.421875, - "learning_rate": 1.1986558033835333e-06, - "loss": 2.5195, - "step": 2844 - }, - { - "epoch": 0.911128903122498, - "grad_norm": 0.427734375, - "learning_rate": 1.190116847870293e-06, - "loss": 2.6438, - "step": 2845 - }, - { - "epoch": 0.911449159327462, - "grad_norm": 1.0, - "learning_rate": 1.181607674331439e-06, - "loss": 2.0309, - "step": 2846 - }, - { - "epoch": 0.911769415532426, - "grad_norm": 0.6015625, - "learning_rate": 1.1731282934104553e-06, - "loss": 2.1805, - "step": 2847 - }, - { - "epoch": 0.9120896717373899, - "grad_norm": 0.48046875, - "learning_rate": 1.1646787157135463e-06, - "loss": 2.2949, - "step": 2848 - }, - { - "epoch": 0.9124099279423539, - "grad_norm": 0.443359375, - "learning_rate": 1.1562589518096607e-06, - "loss": 2.3779, - "step": 2849 - }, - { - "epoch": 0.9127301841473179, - "grad_norm": 0.41015625, - "learning_rate": 1.1478690122304293e-06, - "loss": 1.8918, - "step": 2850 - }, - { - "epoch": 0.9130504403522818, - "grad_norm": 0.6796875, - "learning_rate": 1.1395089074701936e-06, - "loss": 2.4792, - "step": 2851 - }, - { - "epoch": 0.9133706965572458, - "grad_norm": 0.490234375, - "learning_rate": 1.1311786479859781e-06, - "loss": 2.6007, - "step": 2852 - }, - { - "epoch": 0.9136909527622098, - "grad_norm": 0.361328125, - "learning_rate": 1.12287824419747e-06, - "loss": 2.0799, - "step": 2853 - }, - { - "epoch": 0.9140112089671737, - "grad_norm": 0.498046875, - "learning_rate": 1.114607706487006e-06, - "loss": 2.4274, - "step": 2854 - }, - { - "epoch": 0.9143314651721377, - "grad_norm": 0.62109375, - "learning_rate": 1.1063670451995856e-06, - "loss": 2.5174, - "step": 2855 - }, - { - "epoch": 0.9146517213771017, - "grad_norm": 0.52734375, - "learning_rate": 1.0981562706428166e-06, - "loss": 2.1216, - "step": 2856 - }, - { - "epoch": 0.9149719775820656, - "grad_norm": 0.6953125, - "learning_rate": 1.0899753930869394e-06, - "loss": 2.0923, - "step": 2857 - }, - { - "epoch": 0.9152922337870296, - "grad_norm": 0.427734375, - "learning_rate": 1.0818244227647905e-06, - "loss": 2.3559, - "step": 2858 - }, - { - "epoch": 0.9156124899919936, - "grad_norm": 2.15625, - "learning_rate": 1.0737033698717979e-06, - "loss": 2.3995, - "step": 2859 - }, - { - "epoch": 0.9159327461969575, - "grad_norm": 0.48046875, - "learning_rate": 1.0656122445659721e-06, - "loss": 2.1872, - "step": 2860 - }, - { - "epoch": 0.9162530024019215, - "grad_norm": 0.47265625, - "learning_rate": 1.057551056967887e-06, - "loss": 1.9036, - "step": 2861 - }, - { - "epoch": 0.9165732586068855, - "grad_norm": 0.458984375, - "learning_rate": 1.0495198171606657e-06, - "loss": 2.4604, - "step": 2862 - }, - { - "epoch": 0.9168935148118494, - "grad_norm": 0.60546875, - "learning_rate": 1.0415185351899837e-06, - "loss": 2.326, - "step": 2863 - }, - { - "epoch": 0.9172137710168135, - "grad_norm": 0.46484375, - "learning_rate": 1.0335472210640323e-06, - "loss": 2.0918, - "step": 2864 - }, - { - "epoch": 0.9175340272217775, - "grad_norm": 0.59765625, - "learning_rate": 1.0256058847535221e-06, - "loss": 2.2654, - "step": 2865 - }, - { - "epoch": 0.9178542834267414, - "grad_norm": 0.4765625, - "learning_rate": 1.0176945361916657e-06, - "loss": 2.1514, - "step": 2866 - }, - { - "epoch": 0.9181745396317054, - "grad_norm": 0.421875, - "learning_rate": 1.0098131852741777e-06, - "loss": 2.1784, - "step": 2867 - }, - { - "epoch": 0.9184947958366694, - "grad_norm": 0.6171875, - "learning_rate": 1.001961841859228e-06, - "loss": 2.1977, - "step": 2868 - }, - { - "epoch": 0.9188150520416333, - "grad_norm": 0.515625, - "learning_rate": 9.941405157674778e-07, - "loss": 2.4212, - "step": 2869 - }, - { - "epoch": 0.9191353082465973, - "grad_norm": 0.51171875, - "learning_rate": 9.863492167820232e-07, - "loss": 1.8568, - "step": 2870 - }, - { - "epoch": 0.9194555644515613, - "grad_norm": 0.43359375, - "learning_rate": 9.785879546484077e-07, - "loss": 2.2068, - "step": 2871 - }, - { - "epoch": 0.9197758206565252, - "grad_norm": 0.48046875, - "learning_rate": 9.708567390746077e-07, - "loss": 2.2889, - "step": 2872 - }, - { - "epoch": 0.9200960768614892, - "grad_norm": 0.55078125, - "learning_rate": 9.631555797310122e-07, - "loss": 2.3926, - "step": 2873 - }, - { - "epoch": 0.9204163330664532, - "grad_norm": 0.490234375, - "learning_rate": 9.554844862504158e-07, - "loss": 2.2207, - "step": 2874 - }, - { - "epoch": 0.9207365892714171, - "grad_norm": 0.478515625, - "learning_rate": 9.47843468228013e-07, - "loss": 2.3968, - "step": 2875 - }, - { - "epoch": 0.9210568454763811, - "grad_norm": 0.4921875, - "learning_rate": 9.402325352213582e-07, - "loss": 2.2347, - "step": 2876 - }, - { - "epoch": 0.9213771016813451, - "grad_norm": 0.466796875, - "learning_rate": 9.326516967504056e-07, - "loss": 2.4828, - "step": 2877 - }, - { - "epoch": 0.921697357886309, - "grad_norm": 0.53515625, - "learning_rate": 9.251009622974426e-07, - "loss": 2.2806, - "step": 2878 - }, - { - "epoch": 0.922017614091273, - "grad_norm": 0.3828125, - "learning_rate": 9.175803413071088e-07, - "loss": 2.2707, - "step": 2879 - }, - { - "epoch": 0.922337870296237, - "grad_norm": 1.3515625, - "learning_rate": 9.100898431863791e-07, - "loss": 1.9029, - "step": 2880 - }, - { - "epoch": 0.9226581265012009, - "grad_norm": 0.3828125, - "learning_rate": 9.026294773045535e-07, - "loss": 1.0364, - "step": 2881 - }, - { - "epoch": 0.922978382706165, - "grad_norm": 0.56640625, - "learning_rate": 8.95199252993223e-07, - "loss": 2.0725, - "step": 2882 - }, - { - "epoch": 0.923298638911129, - "grad_norm": 0.5859375, - "learning_rate": 8.877991795463086e-07, - "loss": 2.2058, - "step": 2883 - }, - { - "epoch": 0.9236188951160929, - "grad_norm": 0.5078125, - "learning_rate": 8.804292662199842e-07, - "loss": 2.1161, - "step": 2884 - }, - { - "epoch": 0.9239391513210569, - "grad_norm": 0.8828125, - "learning_rate": 8.730895222327284e-07, - "loss": 2.715, - "step": 2885 - }, - { - "epoch": 0.9242594075260209, - "grad_norm": 0.29296875, - "learning_rate": 8.657799567652614e-07, - "loss": 1.7732, - "step": 2886 - }, - { - "epoch": 0.9245796637309848, - "grad_norm": 0.462890625, - "learning_rate": 8.585005789605638e-07, - "loss": 2.3266, - "step": 2887 - }, - { - "epoch": 0.9248999199359488, - "grad_norm": 0.4375, - "learning_rate": 8.512513979238579e-07, - "loss": 2.277, - "step": 2888 - }, - { - "epoch": 0.9252201761409127, - "grad_norm": 0.50390625, - "learning_rate": 8.440324227225959e-07, - "loss": 2.2216, - "step": 2889 - }, - { - "epoch": 0.9255404323458767, - "grad_norm": 0.34375, - "learning_rate": 8.368436623864384e-07, - "loss": 2.1062, - "step": 2890 - }, - { - "epoch": 0.9258606885508407, - "grad_norm": 0.703125, - "learning_rate": 8.296851259072702e-07, - "loss": 2.6251, - "step": 2891 - }, - { - "epoch": 0.9261809447558046, - "grad_norm": 0.5703125, - "learning_rate": 8.225568222391539e-07, - "loss": 3.0058, - "step": 2892 - }, - { - "epoch": 0.9265012009607686, - "grad_norm": 0.46875, - "learning_rate": 8.154587602983432e-07, - "loss": 2.0105, - "step": 2893 - }, - { - "epoch": 0.9268214571657326, - "grad_norm": 0.478515625, - "learning_rate": 8.08390948963264e-07, - "loss": 2.0824, - "step": 2894 - }, - { - "epoch": 0.9271417133706965, - "grad_norm": 0.4921875, - "learning_rate": 8.013533970745113e-07, - "loss": 2.2465, - "step": 2895 - }, - { - "epoch": 0.9274619695756605, - "grad_norm": 0.7265625, - "learning_rate": 7.943461134348129e-07, - "loss": 2.2219, - "step": 2896 - }, - { - "epoch": 0.9277822257806245, - "grad_norm": 0.58203125, - "learning_rate": 7.873691068090605e-07, - "loss": 2.332, - "step": 2897 - }, - { - "epoch": 0.9281024819855884, - "grad_norm": 0.45703125, - "learning_rate": 7.804223859242482e-07, - "loss": 2.7178, - "step": 2898 - }, - { - "epoch": 0.9284227381905524, - "grad_norm": 0.515625, - "learning_rate": 7.735059594695171e-07, - "loss": 1.4854, - "step": 2899 - }, - { - "epoch": 0.9287429943955164, - "grad_norm": 0.51953125, - "learning_rate": 7.666198360960858e-07, - "loss": 2.2354, - "step": 2900 - }, - { - "epoch": 0.9290632506004803, - "grad_norm": 0.462890625, - "learning_rate": 7.597640244172921e-07, - "loss": 1.8512, - "step": 2901 - }, - { - "epoch": 0.9293835068054443, - "grad_norm": 0.5078125, - "learning_rate": 7.529385330085431e-07, - "loss": 1.8892, - "step": 2902 - }, - { - "epoch": 0.9297037630104084, - "grad_norm": 0.38671875, - "learning_rate": 7.461433704073373e-07, - "loss": 2.5126, - "step": 2903 - }, - { - "epoch": 0.9300240192153723, - "grad_norm": 0.9296875, - "learning_rate": 7.393785451132146e-07, - "loss": 2.1277, - "step": 2904 - }, - { - "epoch": 0.9303442754203363, - "grad_norm": 0.453125, - "learning_rate": 7.326440655877925e-07, - "loss": 2.4802, - "step": 2905 - }, - { - "epoch": 0.9306645316253003, - "grad_norm": 0.65234375, - "learning_rate": 7.259399402547134e-07, - "loss": 2.1335, - "step": 2906 - }, - { - "epoch": 0.9309847878302642, - "grad_norm": 1.46875, - "learning_rate": 7.192661774996584e-07, - "loss": 2.5345, - "step": 2907 - }, - { - "epoch": 0.9313050440352282, - "grad_norm": 0.640625, - "learning_rate": 7.126227856703332e-07, - "loss": 2.3705, - "step": 2908 - }, - { - "epoch": 0.9316253002401922, - "grad_norm": 0.609375, - "learning_rate": 7.060097730764548e-07, - "loss": 2.6065, - "step": 2909 - }, - { - "epoch": 0.9319455564451561, - "grad_norm": 0.453125, - "learning_rate": 6.994271479897314e-07, - "loss": 2.1943, - "step": 2910 - }, - { - "epoch": 0.9322658126501201, - "grad_norm": 0.47265625, - "learning_rate": 6.928749186438738e-07, - "loss": 1.9554, - "step": 2911 - }, - { - "epoch": 0.9325860688550841, - "grad_norm": 0.494140625, - "learning_rate": 6.863530932345625e-07, - "loss": 2.3002, - "step": 2912 - }, - { - "epoch": 0.932906325060048, - "grad_norm": 0.59765625, - "learning_rate": 6.798616799194634e-07, - "loss": 2.4129, - "step": 2913 - }, - { - "epoch": 0.933226581265012, - "grad_norm": 0.57421875, - "learning_rate": 6.734006868181847e-07, - "loss": 2.1435, - "step": 2914 - }, - { - "epoch": 0.933546837469976, - "grad_norm": 0.4921875, - "learning_rate": 6.66970122012292e-07, - "loss": 1.9209, - "step": 2915 - }, - { - "epoch": 0.9338670936749399, - "grad_norm": 0.515625, - "learning_rate": 6.605699935452902e-07, - "loss": 2.4338, - "step": 2916 - }, - { - "epoch": 0.9341873498799039, - "grad_norm": 0.44140625, - "learning_rate": 6.542003094226201e-07, - "loss": 2.2675, - "step": 2917 - }, - { - "epoch": 0.9345076060848679, - "grad_norm": 0.431640625, - "learning_rate": 6.478610776116278e-07, - "loss": 1.4878, - "step": 2918 - }, - { - "epoch": 0.9348278622898318, - "grad_norm": 0.6796875, - "learning_rate": 6.415523060415846e-07, - "loss": 2.2072, - "step": 2919 - }, - { - "epoch": 0.9351481184947958, - "grad_norm": 0.458984375, - "learning_rate": 6.352740026036475e-07, - "loss": 2.3663, - "step": 2920 - }, - { - "epoch": 0.9354683746997599, - "grad_norm": 0.5234375, - "learning_rate": 6.290261751508708e-07, - "loss": 2.3013, - "step": 2921 - }, - { - "epoch": 0.9357886309047238, - "grad_norm": 0.5703125, - "learning_rate": 6.228088314981867e-07, - "loss": 2.0383, - "step": 2922 - }, - { - "epoch": 0.9361088871096878, - "grad_norm": 0.53515625, - "learning_rate": 6.16621979422402e-07, - "loss": 1.8135, - "step": 2923 - }, - { - "epoch": 0.9364291433146518, - "grad_norm": 0.94140625, - "learning_rate": 6.104656266621766e-07, - "loss": 2.2271, - "step": 2924 - }, - { - "epoch": 0.9367493995196157, - "grad_norm": 0.5546875, - "learning_rate": 6.043397809180257e-07, - "loss": 2.2725, - "step": 2925 - }, - { - "epoch": 0.9370696557245797, - "grad_norm": 0.58984375, - "learning_rate": 5.982444498523005e-07, - "loss": 2.1763, - "step": 2926 - }, - { - "epoch": 0.9373899119295437, - "grad_norm": 0.5, - "learning_rate": 5.921796410891994e-07, - "loss": 2.3399, - "step": 2927 - }, - { - "epoch": 0.9377101681345076, - "grad_norm": 0.474609375, - "learning_rate": 5.861453622147157e-07, - "loss": 1.8565, - "step": 2928 - }, - { - "epoch": 0.9380304243394716, - "grad_norm": 0.57421875, - "learning_rate": 5.801416207766808e-07, - "loss": 2.3172, - "step": 2929 - }, - { - "epoch": 0.9383506805444356, - "grad_norm": 1.4453125, - "learning_rate": 5.741684242847156e-07, - "loss": 1.864, - "step": 2930 - }, - { - "epoch": 0.9386709367493995, - "grad_norm": 0.53125, - "learning_rate": 5.682257802102408e-07, - "loss": 2.4591, - "step": 2931 - }, - { - "epoch": 0.9389911929543635, - "grad_norm": 0.498046875, - "learning_rate": 5.623136959864494e-07, - "loss": 2.759, - "step": 2932 - }, - { - "epoch": 0.9393114491593275, - "grad_norm": 1.703125, - "learning_rate": 5.564321790083316e-07, - "loss": 1.5265, - "step": 2933 - }, - { - "epoch": 0.9396317053642914, - "grad_norm": 0.54296875, - "learning_rate": 5.505812366326219e-07, - "loss": 2.2601, - "step": 2934 - }, - { - "epoch": 0.9399519615692554, - "grad_norm": 0.50390625, - "learning_rate": 5.447608761778189e-07, - "loss": 1.9696, - "step": 2935 - }, - { - "epoch": 0.9402722177742193, - "grad_norm": 0.423828125, - "learning_rate": 5.389711049241741e-07, - "loss": 2.2893, - "step": 2936 - }, - { - "epoch": 0.9405924739791833, - "grad_norm": 0.515625, - "learning_rate": 5.332119301136695e-07, - "loss": 2.2355, - "step": 2937 - }, - { - "epoch": 0.9409127301841473, - "grad_norm": 0.5859375, - "learning_rate": 5.274833589500205e-07, - "loss": 2.0032, - "step": 2938 - }, - { - "epoch": 0.9412329863891112, - "grad_norm": 0.451171875, - "learning_rate": 5.217853985986648e-07, - "loss": 2.6002, - "step": 2939 - }, - { - "epoch": 0.9415532425940752, - "grad_norm": 0.443359375, - "learning_rate": 5.161180561867401e-07, - "loss": 2.2874, - "step": 2940 - }, - { - "epoch": 0.9418734987990393, - "grad_norm": 0.49609375, - "learning_rate": 5.104813388031038e-07, - "loss": 2.5509, - "step": 2941 - }, - { - "epoch": 0.9421937550040032, - "grad_norm": 0.65625, - "learning_rate": 5.048752534982909e-07, - "loss": 2.1587, - "step": 2942 - }, - { - "epoch": 0.9425140112089672, - "grad_norm": 0.451171875, - "learning_rate": 4.992998072845311e-07, - "loss": 2.4545, - "step": 2943 - }, - { - "epoch": 0.9428342674139312, - "grad_norm": 0.443359375, - "learning_rate": 4.937550071357239e-07, - "loss": 1.585, - "step": 2944 - }, - { - "epoch": 0.9431545236188951, - "grad_norm": 0.5390625, - "learning_rate": 4.882408599874433e-07, - "loss": 2.0142, - "step": 2945 - }, - { - "epoch": 0.9434747798238591, - "grad_norm": 0.5703125, - "learning_rate": 4.827573727369112e-07, - "loss": 2.179, - "step": 2946 - }, - { - "epoch": 0.9437950360288231, - "grad_norm": 0.7578125, - "learning_rate": 4.773045522430103e-07, - "loss": 2.4145, - "step": 2947 - }, - { - "epoch": 0.944115292233787, - "grad_norm": 0.486328125, - "learning_rate": 4.7188240532625714e-07, - "loss": 2.4136, - "step": 2948 - }, - { - "epoch": 0.944435548438751, - "grad_norm": 0.51171875, - "learning_rate": 4.664909387688043e-07, - "loss": 1.7031, - "step": 2949 - }, - { - "epoch": 0.944755804643715, - "grad_norm": 0.4921875, - "learning_rate": 4.6113015931442683e-07, - "loss": 2.3827, - "step": 2950 - }, - { - "epoch": 0.9450760608486789, - "grad_norm": 0.462890625, - "learning_rate": 4.558000736685164e-07, - "loss": 2.4576, - "step": 2951 - }, - { - "epoch": 0.9453963170536429, - "grad_norm": 0.494140625, - "learning_rate": 4.5050068849807334e-07, - "loss": 2.6722, - "step": 2952 - }, - { - "epoch": 0.9457165732586069, - "grad_norm": 0.48046875, - "learning_rate": 4.4523201043169813e-07, - "loss": 2.5232, - "step": 2953 - }, - { - "epoch": 0.9460368294635708, - "grad_norm": 0.474609375, - "learning_rate": 4.3999404605957196e-07, - "loss": 2.327, - "step": 2954 - }, - { - "epoch": 0.9463570856685348, - "grad_norm": 0.5859375, - "learning_rate": 4.347868019334789e-07, - "loss": 2.152, - "step": 2955 - }, - { - "epoch": 0.9466773418734988, - "grad_norm": 0.6015625, - "learning_rate": 4.296102845667532e-07, - "loss": 2.3246, - "step": 2956 - }, - { - "epoch": 0.9469975980784627, - "grad_norm": 0.8984375, - "learning_rate": 4.244645004343156e-07, - "loss": 2.4348, - "step": 2957 - }, - { - "epoch": 0.9473178542834267, - "grad_norm": 0.546875, - "learning_rate": 4.193494559726313e-07, - "loss": 2.3624, - "step": 2958 - }, - { - "epoch": 0.9476381104883908, - "grad_norm": 0.435546875, - "learning_rate": 4.1426515757972684e-07, - "loss": 2.3969, - "step": 2959 - }, - { - "epoch": 0.9479583666933546, - "grad_norm": 1.015625, - "learning_rate": 4.0921161161515674e-07, - "loss": 2.1616, - "step": 2960 - }, - { - "epoch": 0.9482786228983187, - "grad_norm": 0.34765625, - "learning_rate": 4.041888244000286e-07, - "loss": 1.9173, - "step": 2961 - }, - { - "epoch": 0.9485988791032827, - "grad_norm": 0.52734375, - "learning_rate": 3.991968022169529e-07, - "loss": 2.3724, - "step": 2962 - }, - { - "epoch": 0.9489191353082466, - "grad_norm": 0.380859375, - "learning_rate": 3.9423555131007925e-07, - "loss": 2.1529, - "step": 2963 - }, - { - "epoch": 0.9492393915132106, - "grad_norm": 0.51953125, - "learning_rate": 3.8930507788506046e-07, - "loss": 2.1868, - "step": 2964 - }, - { - "epoch": 0.9495596477181746, - "grad_norm": 0.4921875, - "learning_rate": 3.8440538810904367e-07, - "loss": 1.8231, - "step": 2965 - }, - { - "epoch": 0.9498799039231385, - "grad_norm": 0.6171875, - "learning_rate": 3.7953648811068777e-07, - "loss": 2.235, - "step": 2966 - }, - { - "epoch": 0.9502001601281025, - "grad_norm": 0.96484375, - "learning_rate": 3.746983839801238e-07, - "loss": 2.3122, - "step": 2967 - }, - { - "epoch": 0.9505204163330665, - "grad_norm": 0.50390625, - "learning_rate": 3.698910817689694e-07, - "loss": 2.4373, - "step": 2968 - }, - { - "epoch": 0.9508406725380304, - "grad_norm": 0.416015625, - "learning_rate": 3.6511458749031723e-07, - "loss": 1.7601, - "step": 2969 - }, - { - "epoch": 0.9511609287429944, - "grad_norm": 0.7734375, - "learning_rate": 3.6036890711871886e-07, - "loss": 2.7486, - "step": 2970 - }, - { - "epoch": 0.9514811849479584, - "grad_norm": 0.4296875, - "learning_rate": 3.556540465901842e-07, - "loss": 2.1883, - "step": 2971 - }, - { - "epoch": 0.9518014411529223, - "grad_norm": 0.57421875, - "learning_rate": 3.509700118021736e-07, - "loss": 2.4816, - "step": 2972 - }, - { - "epoch": 0.9521216973578863, - "grad_norm": 0.6484375, - "learning_rate": 3.463168086135976e-07, - "loss": 2.3101, - "step": 2973 - }, - { - "epoch": 0.9524419535628503, - "grad_norm": 0.55859375, - "learning_rate": 3.416944428447838e-07, - "loss": 2.7864, - "step": 2974 - }, - { - "epoch": 0.9527622097678142, - "grad_norm": 0.58203125, - "learning_rate": 3.3710292027750724e-07, - "loss": 2.1871, - "step": 2975 - }, - { - "epoch": 0.9530824659727782, - "grad_norm": 0.55078125, - "learning_rate": 3.3254224665494883e-07, - "loss": 2.508, - "step": 2976 - }, - { - "epoch": 0.9534027221777422, - "grad_norm": 0.65234375, - "learning_rate": 3.2801242768171483e-07, - "loss": 1.6672, - "step": 2977 - }, - { - "epoch": 0.9537229783827061, - "grad_norm": 0.451171875, - "learning_rate": 3.235134690238062e-07, - "loss": 2.5607, - "step": 2978 - }, - { - "epoch": 0.9540432345876702, - "grad_norm": 0.5703125, - "learning_rate": 3.1904537630862984e-07, - "loss": 2.4434, - "step": 2979 - }, - { - "epoch": 0.9543634907926342, - "grad_norm": 0.56640625, - "learning_rate": 3.146081551249874e-07, - "loss": 2.1326, - "step": 2980 - }, - { - "epoch": 0.9546837469975981, - "grad_norm": 0.6796875, - "learning_rate": 3.102018110230587e-07, - "loss": 2.3844, - "step": 2981 - }, - { - "epoch": 0.9550040032025621, - "grad_norm": 0.3828125, - "learning_rate": 3.058263495144015e-07, - "loss": 2.5421, - "step": 2982 - }, - { - "epoch": 0.955324259407526, - "grad_norm": 0.46484375, - "learning_rate": 3.014817760719546e-07, - "loss": 1.6864, - "step": 2983 - }, - { - "epoch": 0.95564451561249, - "grad_norm": 0.55078125, - "learning_rate": 2.971680961300127e-07, - "loss": 2.5311, - "step": 2984 - }, - { - "epoch": 0.955964771817454, - "grad_norm": 0.6875, - "learning_rate": 2.9288531508422644e-07, - "loss": 2.2674, - "step": 2985 - }, - { - "epoch": 0.9562850280224179, - "grad_norm": 0.8828125, - "learning_rate": 2.886334382916078e-07, - "loss": 2.4979, - "step": 2986 - }, - { - "epoch": 0.9566052842273819, - "grad_norm": 0.51171875, - "learning_rate": 2.8441247107050264e-07, - "loss": 2.3117, - "step": 2987 - }, - { - "epoch": 0.9569255404323459, - "grad_norm": 0.546875, - "learning_rate": 2.802224187005986e-07, - "loss": 2.069, - "step": 2988 - }, - { - "epoch": 0.9572457966373098, - "grad_norm": 1.1484375, - "learning_rate": 2.7606328642291736e-07, - "loss": 2.3921, - "step": 2989 - }, - { - "epoch": 0.9575660528422738, - "grad_norm": 0.52734375, - "learning_rate": 2.7193507943979457e-07, - "loss": 2.7724, - "step": 2990 - }, - { - "epoch": 0.9578863090472378, - "grad_norm": 0.51953125, - "learning_rate": 2.678378029148998e-07, - "loss": 2.1656, - "step": 2991 - }, - { - "epoch": 0.9582065652522017, - "grad_norm": 0.51171875, - "learning_rate": 2.6377146197319733e-07, - "loss": 2.1326, - "step": 2992 - }, - { - "epoch": 0.9585268214571657, - "grad_norm": 0.66015625, - "learning_rate": 2.5973606170096585e-07, - "loss": 2.2594, - "step": 2993 - }, - { - "epoch": 0.9588470776621297, - "grad_norm": 0.671875, - "learning_rate": 2.5573160714578436e-07, - "loss": 2.2226, - "step": 2994 - }, - { - "epoch": 0.9591673338670936, - "grad_norm": 0.44921875, - "learning_rate": 2.517581033165184e-07, - "loss": 2.1275, - "step": 2995 - }, - { - "epoch": 0.9594875900720576, - "grad_norm": 0.5234375, - "learning_rate": 2.4781555518332e-07, - "loss": 2.3456, - "step": 2996 - }, - { - "epoch": 0.9598078462770216, - "grad_norm": 0.6484375, - "learning_rate": 2.439039676776306e-07, - "loss": 2.5249, - "step": 2997 - }, - { - "epoch": 0.9601281024819855, - "grad_norm": 0.373046875, - "learning_rate": 2.4002334569215023e-07, - "loss": 1.8815, - "step": 2998 - }, - { - "epoch": 0.9604483586869496, - "grad_norm": 0.5546875, - "learning_rate": 2.3617369408085732e-07, - "loss": 2.5048, - "step": 2999 - }, - { - "epoch": 0.9607686148919136, - "grad_norm": 0.53125, - "learning_rate": 2.3235501765898615e-07, - "loss": 2.205, - "step": 3000 - }, - { - "epoch": 0.9610888710968775, - "grad_norm": 0.65625, - "learning_rate": 2.285673212030326e-07, - "loss": 2.0211, - "step": 3001 - }, - { - "epoch": 0.9614091273018415, - "grad_norm": 0.50390625, - "learning_rate": 2.2481060945073738e-07, - "loss": 2.1957, - "step": 3002 - }, - { - "epoch": 0.9617293835068055, - "grad_norm": 0.498046875, - "learning_rate": 2.21084887101089e-07, - "loss": 2.2572, - "step": 3003 - }, - { - "epoch": 0.9620496397117694, - "grad_norm": 0.51953125, - "learning_rate": 2.1739015881430125e-07, - "loss": 2.4002, - "step": 3004 - }, - { - "epoch": 0.9623698959167334, - "grad_norm": 0.5703125, - "learning_rate": 2.137264292118385e-07, - "loss": 1.8485, - "step": 3005 - }, - { - "epoch": 0.9626901521216974, - "grad_norm": 2.34375, - "learning_rate": 2.1009370287637664e-07, - "loss": 2.4418, - "step": 3006 - }, - { - "epoch": 0.9630104083266613, - "grad_norm": 0.5546875, - "learning_rate": 2.0649198435181983e-07, - "loss": 2.4171, - "step": 3007 - }, - { - "epoch": 0.9633306645316253, - "grad_norm": 0.4609375, - "learning_rate": 2.02921278143281e-07, - "loss": 2.5708, - "step": 3008 - }, - { - "epoch": 0.9636509207365893, - "grad_norm": 0.44140625, - "learning_rate": 1.99381588717093e-07, - "loss": 2.2956, - "step": 3009 - }, - { - "epoch": 0.9639711769415532, - "grad_norm": 0.734375, - "learning_rate": 1.9587292050077255e-07, - "loss": 2.6983, - "step": 3010 - }, - { - "epoch": 0.9642914331465172, - "grad_norm": 0.55859375, - "learning_rate": 1.9239527788305622e-07, - "loss": 2.1624, - "step": 3011 - }, - { - "epoch": 0.9646116893514812, - "grad_norm": 0.7265625, - "learning_rate": 1.889486652138589e-07, - "loss": 1.8433, - "step": 3012 - }, - { - "epoch": 0.9649319455564451, - "grad_norm": 0.58203125, - "learning_rate": 1.855330868042876e-07, - "loss": 2.5753, - "step": 3013 - }, - { - "epoch": 0.9652522017614091, - "grad_norm": 0.43359375, - "learning_rate": 1.8214854692662765e-07, - "loss": 1.9499, - "step": 3014 - }, - { - "epoch": 0.9655724579663731, - "grad_norm": 0.54296875, - "learning_rate": 1.7879504981434536e-07, - "loss": 1.9731, - "step": 3015 - }, - { - "epoch": 0.965892714171337, - "grad_norm": 0.466796875, - "learning_rate": 1.7547259966207708e-07, - "loss": 2.0553, - "step": 3016 - }, - { - "epoch": 0.966212970376301, - "grad_norm": 0.51953125, - "learning_rate": 1.7218120062562071e-07, - "loss": 2.3178, - "step": 3017 - }, - { - "epoch": 0.9665332265812651, - "grad_norm": 0.48828125, - "learning_rate": 1.6892085682193583e-07, - "loss": 1.948, - "step": 3018 - }, - { - "epoch": 0.966853482786229, - "grad_norm": 0.58984375, - "learning_rate": 1.6569157232914079e-07, - "loss": 2.2639, - "step": 3019 - }, - { - "epoch": 0.967173738991193, - "grad_norm": 0.365234375, - "learning_rate": 1.6249335118650456e-07, - "loss": 2.1195, - "step": 3020 - }, - { - "epoch": 0.967493995196157, - "grad_norm": 0.54296875, - "learning_rate": 1.5932619739443543e-07, - "loss": 2.0017, - "step": 3021 - }, - { - "epoch": 0.9678142514011209, - "grad_norm": 0.443359375, - "learning_rate": 1.5619011491448677e-07, - "loss": 1.485, - "step": 3022 - }, - { - "epoch": 0.9681345076060849, - "grad_norm": 0.4453125, - "learning_rate": 1.5308510766934848e-07, - "loss": 2.2819, - "step": 3023 - }, - { - "epoch": 0.9684547638110489, - "grad_norm": 0.4375, - "learning_rate": 1.5001117954283329e-07, - "loss": 1.7135, - "step": 3024 - }, - { - "epoch": 0.9687750200160128, - "grad_norm": 0.462890625, - "learning_rate": 1.4696833437988778e-07, - "loss": 2.3383, - "step": 3025 - }, - { - "epoch": 0.9690952762209768, - "grad_norm": 0.921875, - "learning_rate": 1.4395657598657296e-07, - "loss": 1.7648, - "step": 3026 - }, - { - "epoch": 0.9694155324259408, - "grad_norm": 0.5234375, - "learning_rate": 1.4097590813007545e-07, - "loss": 2.0177, - "step": 3027 - }, - { - "epoch": 0.9697357886309047, - "grad_norm": 0.462890625, - "learning_rate": 1.380263345386795e-07, - "loss": 1.8389, - "step": 3028 - }, - { - "epoch": 0.9700560448358687, - "grad_norm": 0.5234375, - "learning_rate": 1.35107858901784e-07, - "loss": 2.0451, - "step": 3029 - }, - { - "epoch": 0.9703763010408326, - "grad_norm": 0.62890625, - "learning_rate": 1.3222048486989102e-07, - "loss": 2.4008, - "step": 3030 - }, - { - "epoch": 0.9706965572457966, - "grad_norm": 0.484375, - "learning_rate": 1.2936421605459494e-07, - "loss": 2.2855, - "step": 3031 - }, - { - "epoch": 0.9710168134507606, - "grad_norm": 0.54296875, - "learning_rate": 1.2653905602858507e-07, - "loss": 2.2895, - "step": 3032 - }, - { - "epoch": 0.9713370696557245, - "grad_norm": 0.349609375, - "learning_rate": 1.2374500832564306e-07, - "loss": 2.4505, - "step": 3033 - }, - { - "epoch": 0.9716573258606885, - "grad_norm": 0.54296875, - "learning_rate": 1.209820764406261e-07, - "loss": 2.6746, - "step": 3034 - }, - { - "epoch": 0.9719775820656525, - "grad_norm": 0.640625, - "learning_rate": 1.1825026382947801e-07, - "loss": 2.5954, - "step": 3035 - }, - { - "epoch": 0.9722978382706164, - "grad_norm": 0.71484375, - "learning_rate": 1.1554957390921828e-07, - "loss": 2.7764, - "step": 3036 - }, - { - "epoch": 0.9726180944755805, - "grad_norm": 0.51953125, - "learning_rate": 1.1288001005793358e-07, - "loss": 2.1283, - "step": 3037 - }, - { - "epoch": 0.9729383506805445, - "grad_norm": 0.412109375, - "learning_rate": 1.1024157561477233e-07, - "loss": 2.1614, - "step": 3038 - }, - { - "epoch": 0.9732586068855084, - "grad_norm": 0.5546875, - "learning_rate": 1.0763427387996128e-07, - "loss": 2.4447, - "step": 3039 - }, - { - "epoch": 0.9735788630904724, - "grad_norm": 0.52734375, - "learning_rate": 1.0505810811476947e-07, - "loss": 2.3156, - "step": 3040 - }, - { - "epoch": 0.9738991192954364, - "grad_norm": 0.5234375, - "learning_rate": 1.0251308154153317e-07, - "loss": 2.5195, - "step": 3041 - }, - { - "epoch": 0.9742193755004003, - "grad_norm": 0.640625, - "learning_rate": 9.999919734362539e-08, - "loss": 2.5644, - "step": 3042 - }, - { - "epoch": 0.9745396317053643, - "grad_norm": 0.6328125, - "learning_rate": 9.75164586654781e-08, - "loss": 2.3122, - "step": 3043 - }, - { - "epoch": 0.9748598879103283, - "grad_norm": 0.349609375, - "learning_rate": 9.506486861255715e-08, - "loss": 2.1659, - "step": 3044 - }, - { - "epoch": 0.9751801441152922, - "grad_norm": 0.423828125, - "learning_rate": 9.264443025137626e-08, - "loss": 2.1065, - "step": 3045 - }, - { - "epoch": 0.9755004003202562, - "grad_norm": 1.359375, - "learning_rate": 9.025514660946921e-08, - "loss": 2.4461, - "step": 3046 - }, - { - "epoch": 0.9758206565252202, - "grad_norm": 0.35546875, - "learning_rate": 8.78970206754176e-08, - "loss": 2.4571, - "step": 3047 - }, - { - "epoch": 0.9761409127301841, - "grad_norm": 0.4765625, - "learning_rate": 8.55700553988148e-08, - "loss": 1.889, - "step": 3048 - }, - { - "epoch": 0.9764611689351481, - "grad_norm": 0.421875, - "learning_rate": 8.327425369029085e-08, - "loss": 2.25, - "step": 3049 - }, - { - "epoch": 0.9767814251401121, - "grad_norm": 0.380859375, - "learning_rate": 8.100961842148758e-08, - "loss": 2.4113, - "step": 3050 - }, - { - "epoch": 0.977101681345076, - "grad_norm": 0.546875, - "learning_rate": 7.877615242506408e-08, - "loss": 2.4718, - "step": 3051 - }, - { - "epoch": 0.97742193755004, - "grad_norm": 0.54296875, - "learning_rate": 7.657385849469678e-08, - "loss": 2.5222, - "step": 3052 - }, - { - "epoch": 0.977742193755004, - "grad_norm": 0.55859375, - "learning_rate": 7.44027393850627e-08, - "loss": 2.1825, - "step": 3053 - }, - { - "epoch": 0.9780624499599679, - "grad_norm": 0.484375, - "learning_rate": 7.226279781185341e-08, - "loss": 1.9089, - "step": 3054 - }, - { - "epoch": 0.978382706164932, - "grad_norm": 0.6953125, - "learning_rate": 7.015403645176111e-08, - "loss": 2.3184, - "step": 3055 - }, - { - "epoch": 0.978702962369896, - "grad_norm": 0.486328125, - "learning_rate": 6.80764579424703e-08, - "loss": 1.9132, - "step": 3056 - }, - { - "epoch": 0.9790232185748599, - "grad_norm": 0.921875, - "learning_rate": 6.603006488266616e-08, - "loss": 1.9538, - "step": 3057 - }, - { - "epoch": 0.9793434747798239, - "grad_norm": 0.66796875, - "learning_rate": 6.40148598320317e-08, - "loss": 2.1982, - "step": 3058 - }, - { - "epoch": 0.9796637309847879, - "grad_norm": 0.5390625, - "learning_rate": 6.203084531123115e-08, - "loss": 1.9992, - "step": 3059 - }, - { - "epoch": 0.9799839871897518, - "grad_norm": 0.53515625, - "learning_rate": 6.007802380191552e-08, - "loss": 2.0144, - "step": 3060 - }, - { - "epoch": 0.9803042433947158, - "grad_norm": 0.5703125, - "learning_rate": 5.815639774672532e-08, - "loss": 1.9409, - "step": 3061 - }, - { - "epoch": 0.9806244995996798, - "grad_norm": 0.365234375, - "learning_rate": 5.6265969549271216e-08, - "loss": 2.4769, - "step": 3062 - }, - { - "epoch": 0.9809447558046437, - "grad_norm": 0.37890625, - "learning_rate": 5.44067415741506e-08, - "loss": 2.4554, - "step": 3063 - }, - { - "epoch": 0.9812650120096077, - "grad_norm": 0.54296875, - "learning_rate": 5.257871614692822e-08, - "loss": 2.1183, - "step": 3064 - }, - { - "epoch": 0.9815852682145717, - "grad_norm": 0.75, - "learning_rate": 5.07818955541417e-08, - "loss": 2.0154, - "step": 3065 - }, - { - "epoch": 0.9819055244195356, - "grad_norm": 0.482421875, - "learning_rate": 4.901628204330155e-08, - "loss": 2.1623, - "step": 3066 - }, - { - "epoch": 0.9822257806244996, - "grad_norm": 0.625, - "learning_rate": 4.728187782287452e-08, - "loss": 1.8426, - "step": 3067 - }, - { - "epoch": 0.9825460368294636, - "grad_norm": 0.466796875, - "learning_rate": 4.5578685062297456e-08, - "loss": 2.0737, - "step": 3068 - }, - { - "epoch": 0.9828662930344275, - "grad_norm": 0.62109375, - "learning_rate": 4.390670589196622e-08, - "loss": 2.1993, - "step": 3069 - }, - { - "epoch": 0.9831865492393915, - "grad_norm": 0.9140625, - "learning_rate": 4.2265942403227345e-08, - "loss": 2.5824, - "step": 3070 - }, - { - "epoch": 0.9835068054443555, - "grad_norm": 0.70703125, - "learning_rate": 4.065639664839471e-08, - "loss": 2.3295, - "step": 3071 - }, - { - "epoch": 0.9838270616493194, - "grad_norm": 0.58203125, - "learning_rate": 3.907807064072178e-08, - "loss": 1.9268, - "step": 3072 - }, - { - "epoch": 0.9841473178542834, - "grad_norm": 0.490234375, - "learning_rate": 3.7530966354418215e-08, - "loss": 2.4965, - "step": 3073 - }, - { - "epoch": 0.9844675740592475, - "grad_norm": 0.345703125, - "learning_rate": 3.6015085724638856e-08, - "loss": 2.1137, - "step": 3074 - }, - { - "epoch": 0.9847878302642114, - "grad_norm": 0.326171875, - "learning_rate": 3.453043064748362e-08, - "loss": 1.6508, - "step": 3075 - }, - { - "epoch": 0.9851080864691754, - "grad_norm": 0.478515625, - "learning_rate": 3.307700297999483e-08, - "loss": 2.1851, - "step": 3076 - }, - { - "epoch": 0.9854283426741393, - "grad_norm": 0.314453125, - "learning_rate": 3.1654804540157124e-08, - "loss": 1.754, - "step": 3077 - }, - { - "epoch": 0.9857485988791033, - "grad_norm": 0.37890625, - "learning_rate": 3.02638371068864e-08, - "loss": 2.1616, - "step": 3078 - }, - { - "epoch": 0.9860688550840673, - "grad_norm": 0.53515625, - "learning_rate": 2.890410242003816e-08, - "loss": 2.733, - "step": 3079 - }, - { - "epoch": 0.9863891112890312, - "grad_norm": 0.474609375, - "learning_rate": 2.757560218040467e-08, - "loss": 2.0874, - "step": 3080 - }, - { - "epoch": 0.9867093674939952, - "grad_norm": 0.48828125, - "learning_rate": 2.6278338049706696e-08, - "loss": 2.343, - "step": 3081 - }, - { - "epoch": 0.9870296236989592, - "grad_norm": 0.50390625, - "learning_rate": 2.5012311650587927e-08, - "loss": 2.3688, - "step": 3082 - }, - { - "epoch": 0.9873498799039231, - "grad_norm": 0.47265625, - "learning_rate": 2.377752456662885e-08, - "loss": 1.8563, - "step": 3083 - }, - { - "epoch": 0.9876701361088871, - "grad_norm": 0.53125, - "learning_rate": 2.257397834233288e-08, - "loss": 2.2285, - "step": 3084 - }, - { - "epoch": 0.9879903923138511, - "grad_norm": 0.4921875, - "learning_rate": 2.1401674483118027e-08, - "loss": 2.2832, - "step": 3085 - }, - { - "epoch": 0.988310648518815, - "grad_norm": 0.5625, - "learning_rate": 2.026061445533356e-08, - "loss": 2.1044, - "step": 3086 - }, - { - "epoch": 0.988630904723779, - "grad_norm": 0.58984375, - "learning_rate": 1.9150799686246112e-08, - "loss": 2.234, - "step": 3087 - }, - { - "epoch": 0.988951160928743, - "grad_norm": 0.41796875, - "learning_rate": 1.8072231564036922e-08, - "loss": 1.7073, - "step": 3088 - }, - { - "epoch": 0.9892714171337069, - "grad_norm": 0.490234375, - "learning_rate": 1.702491143780738e-08, - "loss": 2.6741, - "step": 3089 - }, - { - "epoch": 0.9895916733386709, - "grad_norm": 1.1484375, - "learning_rate": 1.6008840617565134e-08, - "loss": 2.2903, - "step": 3090 - }, - { - "epoch": 0.9899119295436349, - "grad_norm": 0.52734375, - "learning_rate": 1.5024020374243554e-08, - "loss": 2.1201, - "step": 3091 - }, - { - "epoch": 0.9902321857485988, - "grad_norm": 0.65234375, - "learning_rate": 1.4070451939673934e-08, - "loss": 2.3257, - "step": 3092 - }, - { - "epoch": 0.9905524419535628, - "grad_norm": 0.39453125, - "learning_rate": 1.3148136506604958e-08, - "loss": 2.3063, - "step": 3093 - }, - { - "epoch": 0.9908726981585269, - "grad_norm": 0.42578125, - "learning_rate": 1.2257075228688797e-08, - "loss": 2.2024, - "step": 3094 - }, - { - "epoch": 0.9911929543634908, - "grad_norm": 0.42578125, - "learning_rate": 1.1397269220486672e-08, - "loss": 2.2727, - "step": 3095 - }, - { - "epoch": 0.9915132105684548, - "grad_norm": 0.80078125, - "learning_rate": 1.0568719557468853e-08, - "loss": 1.8462, - "step": 3096 - }, - { - "epoch": 0.9918334667734188, - "grad_norm": 0.44140625, - "learning_rate": 9.771427276000778e-09, - "loss": 1.2467, - "step": 3097 - }, - { - "epoch": 0.9921537229783827, - "grad_norm": 0.27734375, - "learning_rate": 9.00539337335693e-09, - "loss": 1.7548, - "step": 3098 - }, - { - "epoch": 0.9924739791833467, - "grad_norm": 0.55859375, - "learning_rate": 8.270618807706965e-09, - "loss": 1.8665, - "step": 3099 - }, - { - "epoch": 0.9927942353883107, - "grad_norm": 0.400390625, - "learning_rate": 7.56710449813236e-09, - "loss": 1.8483, - "step": 3100 - }, - { - "epoch": 0.9931144915932746, - "grad_norm": 0.75, - "learning_rate": 6.894851324595886e-09, - "loss": 2.3824, - "step": 3101 - }, - { - "epoch": 0.9934347477982386, - "grad_norm": 0.48828125, - "learning_rate": 6.2538601279776846e-09, - "loss": 2.3081, - "step": 3102 - }, - { - "epoch": 0.9937550040032026, - "grad_norm": 0.53515625, - "learning_rate": 5.644131710039191e-09, - "loss": 2.6313, - "step": 3103 - }, - { - "epoch": 0.9940752602081665, - "grad_norm": 0.42578125, - "learning_rate": 5.065666833442562e-09, - "loss": 2.2497, - "step": 3104 - }, - { - "epoch": 0.9943955164131305, - "grad_norm": 0.478515625, - "learning_rate": 4.518466221750672e-09, - "loss": 2.2095, - "step": 3105 - }, - { - "epoch": 0.9947157726180945, - "grad_norm": 2.53125, - "learning_rate": 4.002530559410467e-09, - "loss": 2.0491, - "step": 3106 - }, - { - "epoch": 0.9950360288230584, - "grad_norm": 0.482421875, - "learning_rate": 3.5178604917668334e-09, - "loss": 1.9228, - "step": 3107 - }, - { - "epoch": 0.9953562850280224, - "grad_norm": 0.427734375, - "learning_rate": 3.0644566250598305e-09, - "loss": 2.6178, - "step": 3108 - }, - { - "epoch": 0.9956765412329864, - "grad_norm": 0.578125, - "learning_rate": 2.6423195264163593e-09, - "loss": 2.0714, - "step": 3109 - }, - { - "epoch": 0.9959967974379503, - "grad_norm": 0.5703125, - "learning_rate": 2.2514497238557144e-09, - "loss": 2.1848, - "step": 3110 - }, - { - "epoch": 0.9963170536429143, - "grad_norm": 0.484375, - "learning_rate": 1.891847706286809e-09, - "loss": 2.4682, - "step": 3111 - }, - { - "epoch": 0.9966373098478783, - "grad_norm": 0.462890625, - "learning_rate": 1.5635139235109507e-09, - "loss": 2.3354, - "step": 3112 - }, - { - "epoch": 0.9969575660528422, - "grad_norm": 0.5546875, - "learning_rate": 1.2664487862107389e-09, - "loss": 1.8637, - "step": 3113 - }, - { - "epoch": 0.9972778222578063, - "grad_norm": 0.408203125, - "learning_rate": 1.0006526659667171e-09, - "loss": 2.0926, - "step": 3114 - }, - { - "epoch": 0.9975980784627703, - "grad_norm": 0.53515625, - "learning_rate": 7.661258952434969e-10, - "loss": 2.0889, - "step": 3115 - }, - { - "epoch": 0.9979183346677342, - "grad_norm": 0.5703125, - "learning_rate": 5.628687673897571e-10, - "loss": 1.7024, - "step": 3116 - }, - { - "epoch": 0.9982385908726982, - "grad_norm": 0.5234375, - "learning_rate": 3.9088153664379457e-10, - "loss": 2.477, - "step": 3117 - }, - { - "epoch": 0.9985588470776622, - "grad_norm": 0.470703125, - "learning_rate": 2.5016441813630054e-10, - "loss": 2.363, - "step": 3118 - }, - { - "epoch": 0.9988791032826261, - "grad_norm": 1.9765625, - "learning_rate": 1.4071758787648214e-10, - "loss": 2.3924, - "step": 3119 - }, - { - "epoch": 0.9991993594875901, - "grad_norm": 1.265625, - "learning_rate": 6.254118276316501e-11, - "loss": 2.2253, - "step": 3120 - }, - { - "epoch": 0.9995196156925541, - "grad_norm": 0.64453125, - "learning_rate": 1.5635300579242008e-11, - "loss": 2.4826, - "step": 3121 - }, - { - "epoch": 0.999839871897518, - "grad_norm": 0.61328125, - "learning_rate": 0.0, - "loss": 2.4466, - "step": 3122 - }, - { - "epoch": 0.999839871897518, - "step": 3122, - "total_flos": 1.7832492586531553e+18, - "train_loss": 2.216206028322468, - "train_runtime": 33247.722, - "train_samples_per_second": 0.751, - "train_steps_per_second": 0.094 + "epoch": 0.9996843204390374, + "step": 2573, + "total_flos": 1.1741952995733012e+19, + "train_loss": 2.4089672912015, + "train_runtime": 234538.8638, + "train_samples_per_second": 1.405, + "train_steps_per_second": 0.011 } ], "logging_steps": 1, - "max_steps": 3122, + "max_steps": 2573, "num_input_tokens_seen": 0, "num_train_epochs": 1, - "save_steps": 100, + "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { @@ -21889,8 +18046,8 @@ "attributes": {} } }, - "total_flos": 1.7832492586531553e+18, - "train_batch_size": 4, + "total_flos": 1.1741952995733012e+19, + "train_batch_size": 8, "trial_name": null, "trial_params": null }